ARCHIPRACTICA

Independent publication. Every tool tested on the same real brief. No sponsored reviews.

P–LIB–01Library

Prompt library for architects

115 prompts · updated October 2026

Every prompt below was extracted from a published guide and run against a real brief before listing. Filter, copy, adapt to your project.

Tool
Task
Stage

P–01ChatGPTConcept

Adjacency notes from a space program

Extracts adjacency requirements from a space program with strength ratings and program quotes, and reports conflicts only where the text proves them.

ChatGPTBrief analysisConcept

From the space program below, produce: 1) a table of required adjacencies -
space pair, strength (must touch / near / no relation / keep apart), and the
reason from the program text; 2) spaces whose placement the program leaves
undecided, as questions for the client; 3) conflicts where the program's
stated adjacencies cannot all hold on one floor, if any, quoting the
conflicting lines. Use only the program text. Program: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–02ChatGPTConcept

Site narrative draft that marks its own gaps

Drafts a client-facing design narrative from site inventory facts, inserting [VERIFY] markers where facts are missing instead of inventing them.

ChatGPTBrief analysisConcept

Draft a one-page design narrative for a landscape project from the site
inventory facts I paste: existing conditions, constraints, program, and the
design moves we chose. Write in plain professional language for a client
audience - no marketing adjectives, no claims beyond the facts I give you.
Where a fact is missing for a claim the narrative needs, insert
[VERIFY: what's missing] instead of inventing it. Inventory and program:
[paste]
Tested with GPT-6 Astra,

From the guide: /guides/chatgpt-for-landscape-architects/

P–03ChatGPTPre-design

Pre-design brief from scattered emails

Builds a structured pre-design brief from scattered client emails and meeting notes, quoting contradictions between sources and ranking open questions by rework risk.

ChatGPTBrief analysisPre-design

I am an architect starting [project type] for [client type]. Below are my raw
notes and excerpts from client emails, separated by ---. Build a structured
pre-design brief under these headings: Project goals / Spaces and requirements /
Budget and schedule signals / Site facts stated / Constraints stated /
Contradictions between sources, quoting both lines / Decisions the client
assumes are made but never confirmed / Open questions, ranked by rework risk.
Use only what appears in my material; where something is implied but not
stated, put it under Open questions. Write [NEEDS INPUT] where a heading has
nothing. Material: ---[paste]---
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–04ChatGPTPractice management

Boilerplate tailoring pass for a new client

Re-orders a standard proposal section around a new client priorities, swaps in the relevant examples and cuts what does not serve them - keeping every factual claim exactly as written.

ChatGPTClient communicationPractice management

Below is a standard section from our past proposals and the priorities of a
new client. Rewrite the section so it speaks to this client's priorities:
reorder points so their top concern comes first, replace generic examples
with the relevant ones from the list I give you, and cut anything that does
not serve this client. Keep every factual claim exactly as written - do not
add projects, numbers, or capabilities. Keep it under [N] words.

Our standard section: [paste]
This client's priorities: [2-4 bullets from the RFP or your calls]
Relevant examples to use: [projects from your portfolio, one line each]
Tested with GPT-6 Astra,

From the guide: /business/chatgpt-architecture-proposals/

P–05ChatGPTPractice management

Project understanding and approach draft

Drafts the understanding-and-approach section of an architecture proposal from your facts pack, mirroring the client vocabulary and inserting [GAP] markers where facts are missing.

ChatGPTClient communicationPractice management

Draft the "Project Understanding and Approach" section of an architecture
proposal, 350-450 words. Use ONLY the facts below - if something important is
missing, insert [GAP: what is missing] instead of inventing it. Mirror the
client's own vocabulary from the RFP excerpt. Structure: what the client is
trying to achieve, the two or three hardest problems in this project, and how
our process addresses each in sequence. Plain professional language, no
superlatives, no "world-class", no "state-of-the-art".

Client and project facts: [paste RFP excerpt or summary]
Our relevant experience: [2-4 comparable projects, one line each]
Our process: [your stage structure, one line per stage]
Tested with GPT-6 Astra,

From the guide: /business/chatgpt-architecture-proposals/

P–06ChatGPTConstruction admin

Change narrative for the client

Writes a client change narrative that opens with the change and its effect, keeps every figure exact and stays neutral.

ChatGPTClient communicationConstruction admin

Write a change narrative for my client from the facts below: what changed,
why, cost and schedule effect as I state them, and what happens next. Open
with the change and its effect in two sentences, then the background.
Neutral tone - no blame, no apology spiral, no minimizing words like "just"
or "small". Keep every figure exactly as I give it. Maximum [250] words.
Facts: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–07ChatGPTPre-design

Kickoff questionnaire for a specific client

Prepares a plain-language kickoff questionnaire grouped by use, budget, schedule, decisions, taste and fears, plus the three questions to ask in person.

ChatGPTClient communicationPre-design

Prepare a kickoff questionnaire for a [project type] client who is [client
profile]. 20 questions maximum, grouped: how they will use the spaces /
budget and its flexibility / schedule drivers / decision-making and approvals
/ taste and references / fears. Phrase every question for someone with no
construction vocabulary, one line each. End with the three questions I should
ask in person rather than in writing, with one line on why.
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–08ChatGPTPre-design

Community input synthesis with honest counts

Groups open-ended public comments into themes with verifiable counts, verbatim examples, and an explicit unthemed list.

ChatGPTClient communicationPre-design

I paste open-ended comments from a community engagement session for a
public landscape project. Group them into themes. For each theme: a name,
how many comments fall into it, two verbatim example comments, and any
direct conflicts between themes. Then list every comment that fits no
theme - do not force them. Keep the count honest: the theme counts must sum
to the comments I pasted minus the unthemed ones. Comments: [paste]
Tested with GPT-6 Astra,

From the guide: /guides/chatgpt-for-landscape-architects/

P–09ChatGPTConstruction admin

Meeting minutes with TBC discipline

Raw site-meeting notes turned into formal minutes, a decisions list and an action table. In our test GPT-6 Astra also caught, unprompted, that a stated 'Fri 24 Oct' fell on a Saturday and marked it TBC.

ChatGPTClient communicationConstruction admin

Turn my raw site-meeting notes below into: 1) formal meeting minutes
(project, date, attendees, items discussed); 2) a decisions list; 3) an
action items table with owner and due date. Use only what is in the notes;
where an owner or date is missing, write TBC rather than inventing one.
Tested with GPT-6 Astra,

From the guide: /tools/chatgpt-vs-gemini-for-architects/

P–10ChatGPTDesign development

Plain-language note on a technical change

Explains a construction change to a client with no technical background in 150 words: why it saves money, what changes visually, what stays the same, one decision.

ChatGPTClient communicationDesign development

Write 150 words maximum that I can send to my client, who has no construction background, explaining this change: [describe the change]. Cover: roughly why it costs less (or more), what changes visually, and what does not change. No technical jargon without a plain-language gloss, no hype, and end with the one decision I need from them.
Tested with GPT-6 Astra,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–11ChatGPTConcept

Plain-language scheme explanation for the client

Rewrites a technical scheme description for a client with no construction background, keeping every number and marking what is still undecided.

ChatGPTClient communicationConcept

Rewrite my technical scheme description for the client, who has no
construction background. Keep every number and dimension unchanged. Replace
jargon with plain equivalents, keeping the technical term in parentheses at
first use. Order: what they will see arriving, how the main spaces work day
to day, what is decided, what is still open and when each open item gets
decided. Maximum [300] words. Description: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–12ChatGPTPractice management

Schedule slip email to the client

Writes a schedule slip email that opens with the revised plan and date, explains the cause plainly and closes with the next step.

ChatGPTClient communicationPractice management

Write an email to my client about a schedule slip. Facts: [what slipped, how
long, why, what continues in parallel, new date]. Open with the revised plan
and new date - not with the problem. Explain the cause in two plain
sentences without blaming anyone. Close with the next concrete step and its
date. Professional, warm, no filler apologies, maximum [180] words. I will
personalize before sending.
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–13ChatGPTDocumentation

Accessibility screen from pasted standard text

Screens a design description against pasted accessibility standard text, with honest appears-to-meet / appears-short / not-covered verdicts.

ChatGPTCode & zoningDocumentation

I paste sections of the accessibility standard adopted in my jurisdiction,
then describe my design. Produce a screening list: requirement with its
section number quoted, what my description says, and "appears to meet",
"appears short", or "not covered by the pasted text". Use only the pasted
sections - no requirements from memory. This prepares my own verification;
the governing check happens on the adopted edition with my code consultant.
Standard: [paste] Design: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–14ChatGPTDocumentation

Code lookup with sections to verify

Egress or other code basics answered with exact section and table numbers, the edition stated, and a list of what to verify in the locally adopted code.

ChatGPTCode & zoningDocumentation

I am an architect checking egress basics for a [building type] in a jurisdiction that has adopted the [code edition] without amendments (I will verify local amendments separately). Answer these, citing the exact section or table number for each: [your questions]. State the code edition you are citing, and list what I must verify in the adopted local code before relying on any of this.
Tested with GPT-6 Astra,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–15ChatGPTDocumentation

Standards retrieval from pasted code text

Answers a question from pasted standard or municipal code text only - quotes the governing section, refuses to fill gaps from memory.

ChatGPTCode & zoningDocumentation

I paste sections of a design standard or municipal code below. Answer my
question using only the pasted text: quote the governing passage with its
section number, state the answer, and say plainly if the pasted text does
not cover my question. Do not fill gaps from memory - if you are using
anything beyond the pasted text, say so explicitly. Standard text: [paste]
Question: [paste]
Tested with GPT-6 Astra,

From the guide: /guides/chatgpt-for-landscape-architects/

P–16ChatGPTPre-design

Buildable envelope worksheet from pasted zoning text

Table every zoning clause with a verbatim quote, derive the maximum footprint and volume with the arithmetic shown, and list what the pasted text leaves unclear.

ChatGPTCode & zoningPre-design

Here is the zoning text that applies to my site, and my lot dimensions. Using
ONLY the text I paste, build a buildable envelope worksheet:

- One row per parameter (setbacks by edge, maximum height, FAR, lot coverage,
  parking minimum, any other limit in the text), with the clause number and a
  verbatim quote of the controlling sentence.
- A derived-envelope section: from those parameters and my lot dimensions,
  calculate the maximum footprint and buildable volume, showing the arithmetic.
- An "unclear from provided text" list for anything the text references but
  does not define (overlays, variances, definitions in other sections).

Do not fill gaps from general knowledge of zoning codes. This is a first-pass
worksheet; a licensed professional verifies it against the current local code.

Zoning text: [paste]
Lot dimensions: [paste]
Tested with GPT-6 Astra,

From the guide: /guides/ai-site-analysis/

P–17ChatGPTPre-design

Zoning parameter table from pasted ordinance text

Organizes pasted zoning ordinance text against proposal parameters in a table with verbatim section quotes - a screen for the architect's own review, never a code check.

ChatGPTCode & zoningPre-design

Below I paste sections of the zoning ordinance for my site, copied from the
adopted text, then my proposal's parameters. Build a table: requirement, the
ordinance line quoted verbatim with its section number, required value,
proposed value, and one of "appears to comply", "appears not to comply",
"cannot tell from the pasted text". Use only the pasted text - do not add
requirements from memory. This organizes my own review; it is not a code
check. Ordinance: [paste] Proposal: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–18ChatGPTPre-design

Restricted-text zoning screen with unclear flags

First-pass zoning screen locked to pasted ordinance text: clause-numbered table with 'appears to meet / appears short / unclear from provided text' flags - tested with a planted violation in both model families.

ChatGPTCode & zoningPre-design

Here is the text of a zoning ordinance section for my site, and my proposed building parameters. Using ONLY the ordinance text I provide, build a table: each requirement with its clause number, my proposed value, and a flag "appears to meet", "appears short" or "unclear from provided text". Do not state that anything complies; this is a first-pass screen. If a requirement cannot be evaluated from the data given, say so.
Tested with GPT-6 Astra and Claude Sonnet 5,

From the guide: /guides/ai-building-code-checks/

P–19ChatGPTConstruction admin

Field report from dictated notes

Turns dictated site notes into a field report with actionable directives, flagging internal inconsistencies in dates and counts instead of resolving them silently.

ChatGPTConstruction admin

Turn my dictated notes from today's site visit into a field report: project,
date, weather, work observed by trade, deficiencies observed, directives
given (imperative sentences a contractor can act on), information requested,
next visit. Keep my wording for deficiencies, cleaned only for clarity. Mark
gaps [NEEDS INPUT] instead of inventing. Flag any internal inconsistency in
my notes - dates, counts, names - rather than resolving it silently. Notes:
[paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–20ChatGPTConstruction admin

Punch list organizer from walkthrough notes

Organizes walkthrough notes into a numbered punch list by room and trade, marking entries too vague to assign.

ChatGPTConstruction admin

Organize my walkthrough notes into a punch list: items grouped by room, then
by trade within each room; each item gets a number, location, deficiency in
one line, and a reference to spec section or drawing if my notes include
one. Keep my deficiency wording. Mark entries too vague to assign [CLARIFY]
- I will re-walk those. End with a count by trade. Notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–21ChatGPTConstruction admin

RFI response draft from pasted documents

Drafts an RFI response from pasted project documents only, citing each document and listing what is missing instead of filling gaps.

ChatGPTConstruction admin

Draft a response to the RFI below using only the project documents I paste
with it. Structure: restate the question in one line; answer it citing each
document by name, sheet or section for every statement; if the pasted
documents do not fully answer it, say exactly what is missing instead of
filling the gap. Neutral wording, no fault assignment. I will review and
sign - note at the end what I must verify before issuing. RFI: [paste]
Documents: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–22ChatGPTConstruction admin

Site-meeting notes to minutes with TBC discipline

Raw site notes become formal minutes, a decisions list and an action table with owner and due date - missing data marked TBC, never invented.

ChatGPTConstruction admin

Turn my raw site-meeting notes below into: 1) formal meeting minutes (project, date, attendees, items discussed); 2) a decisions list; 3) an action items table with owner and due date. Use only what is in the notes; where an owner or date is missing, write TBC rather than inventing one.
Tested with GPT-6 Astra,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–23ChatGPTConstruction admin

Submittal review notes vs spec section

Compares a submittal against the spec section, classifying deviations and omissions as questions for the architect's review action.

ChatGPTConstruction admin

Compare the submittal below against the specification section I paste.
Return: items that match the spec (reference both), deviations (quote both
sides, classify: substitution request / error / clarification needed), and
items the submittal omits that the spec requires. Phrase every deviation as
a question for my review action - the review stamp decision is mine.
Submittal: [paste] Spec section: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–24ChatGPTDesign development

Architectural notes vs consultant report contradictions

Finds contradictions between architectural notes and a consultant report, quoting both sides verbatim and grading severity.

ChatGPTCoordinationDesign development

Below are my architectural notes and the structural engineer's report for
the same project, separated by ===. List only the contradictions between
them: element, what my notes say, what the report says, and severity (stops
coordination / needs a call / minor). Quote both sides verbatim. Do not
propose engineering solutions, and do not flag items that merely appear in
one document and not the other unless they conflict. My notes: [paste] ===
Report: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–25ChatGPTPre-design

Consultant scope question list

Produces per-discipline question tables for consultant fee proposals: scope boundaries, deliverables, exclusions that hurt later, and information owed by phase.

ChatGPTCoordinationPre-design

I am assembling the consultant team for [project type, size]. For each
discipline I list below, produce the questions their fee proposal must
answer: scope boundaries, deliverables by phase, meetings and site visits
included, exclusions that usually hurt later, and information they will need
from me by [phase]. Format: one table per discipline, question / why it
matters. Do not estimate fees. Disciplines: [list]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–26ChatGPTConcept

Concept narrative from design moves

Turns a list of design moves with reasons into a concept narrative for a named audience without adding claims the architect did not make.

ChatGPTDesign narrativesConcept

Turn my design moves into a concept narrative of about [350] words for
[audience]. I list each move with its reason. Keep my reasoning exactly - do
not add sustainability, wellbeing or community claims I did not make, and do
not reorder cause and effect. Plain professional prose, no marketing
adjectives. Moves: 1. [move] because [reason] 2. ...
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–27ChatGPTPractice management

Red-team review against the RFP scoring criteria

Plays the evaluation committee: scores your draft section against the RFP criteria, quotes the weakest sentences and returns the three changes most likely to improve the score.

ChatGPTDesign reviewPractice management

You are the client's evaluation committee. Below are the scoring criteria
from the RFP and our draft response section. Score our draft against each
criterion as the committee would: where does it answer the criterion
directly, where is it generic, and where does it claim something without
evidence? Quote the weakest sentences. Then list the three changes that
would most improve the score. Do not rewrite the section - critique it.

Scoring criteria: [paste from the RFP]
Our draft: [paste the section]
Tested with GPT-6 Astra,

From the guide: /business/chatgpt-architecture-proposals/

P–28ChatGPTConcept

Design options comparison matrix

Builds a design options comparison table strictly from the architect's own descriptions and closes with the numbered decisions the client must make.

ChatGPTDesign reviewConcept

I will present [N] design options for the same program. From my descriptions
below, build a comparison table: how each option answers the program, site
response, structural implications I stated, relative cost drivers I stated,
and risks or unknowns. Use only what I wrote - no invented advantages. After
the table, list the decisions the client must make to choose, as numbered
questions. Options: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–29ChatGPTConcept

Skeptical review rehearsal

Rehearses a design review by generating the hardest jury questions from a scheme summary, ordered by damage, with the gaps that would close them.

ChatGPTDesign reviewConcept

Act as a skeptical design jury reviewing my scheme the day before the real
review. From my summary below, write the eight hardest questions I will
face, ordered from most to least damaging, each with one line naming the
weakness behind it. Then pick the two questions I currently cannot answer
from my own summary and say what material would close each. Do not soften
the questions. Scheme: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–30ChatGPTDesign development

Value-engineering impact questions

Turns contractor value-engineering items into review questions covering performance, durability, appearance, code re-checks and knock-on effects.

ChatGPTDesign reviewDesign development

The contractor proposed the value-engineering items below. For each, list
what it would change beyond cost: performance, durability, appearance,
code-sensitive aspects to re-verify, and knock-on effects on other trades -
each phrased as a question for my review, not a verdict. Do not accept or
reject any item and do not invent savings figures. Items: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–31ChatGPTDesign development

DD issue completeness screen

Screens a design development sheet list for missing content, premature CD-scope material and consultant drawings to confirm.

ChatGPTDocumentation QADesign development

Here is the sheet list of my design development issue for a [project type].
Against the deliverables an architect would normally include at DD for this
building type, list: sheets or content that seem missing, sheets that look
like construction-document scope arriving too early, and the consultant
drawings I should confirm are coming. Format as three lists, one line each
with the reason. My office standard and the contract govern - mark every
item "confirm against contract". Sheet list: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–32ChatGPTDocumentation

General notes duplicate and conflict pass

Passes a drawing set's general notes for duplicates, contradictions and notes that belong in the spec, numbering every finding.

ChatGPTDocumentation QADocumentation

Here are the general notes collected from my drawing set. Find: duplicate
notes saying the same thing in different words (quote both), notes that
contradict each other (quote both), and notes that belong in the
specification rather than on drawings. Suggest a merged wording only for
exact duplicates; leave contradictions for my decision. Number every
finding. Notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–33ChatGPTDesign development

Opening schedule skeleton from markups

Builds a door and window schedule skeleton from markup notes, filling only supported cells and listing openings the notes imply but never describe.

ChatGPTDocumentation QADesign development

Build a door and window schedule skeleton from my markup notes below.
Columns: mark, location, nominal size, type, material, fire or acoustic
rating, hardware group, remarks. Fill only cells my notes support; write
[NEEDS INPUT] in every other cell rather than assuming. After the table,
list openings my notes imply but never describe (a room with no door listed,
a habitable room with no window). Notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–34ChatGPTDocumentation

Permit cover letter and transmittal

Drafts a factual permit cover letter and transmittal, flagging checklist items not yet included as to follow.

ChatGPTDocumentation QADocumentation

Draft a permit submission cover letter and document transmittal from the
facts below: project, address, scope summary, the authority's checklist
items I paste, and which documents I am including. The letter is factual and
brief - no persuasion, no compliance claims beyond "submitted for review".
The transmittal lists each document with title, date and sheet count, and
flags checklist items I have not included as "to follow". Facts: [paste]
Authority checklist: [paste] My documents: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–35ChatGPTDocumentation

Spec section vs drawing notes contradictions

Finds direct contradictions between a spec section and drawing notes, quoted verbatim and classified by pricing impact.

ChatGPTDocumentation QADocumentation

Below are excerpts from my specification section and the notes on my
drawings for the same scope, separated by ===. Find only direct
contradictions: element, what the spec says, what the drawing note says,
quoted verbatim. Classify each: blocks pricing / causes an RFI / cosmetic.
Ignore wording differences that do not change the requirement. Do not
propose which side is right. Spec: [paste] === Drawing notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–36ChatGPTPractice management

Additional services letter

Drafts an additional services letter that maps change requests against the agreement clause, quoted once, with authorization before work.

ChatGPTFees & contractsPractice management

Draft a letter to my client about additional services. I paste: the original
scope summary, the change requests to date with dates, and the agreement
clause on additional services, quoted verbatim. Structure: what the
agreement covers, what has changed, why the new work falls outside, and the
proposed path - written authorization before work proceeds. Collaborative
but firm; cite the clause exactly once. Maximum [350] words. Facts: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–37ChatGPTPractice management

Go/no-go triage for an architecture RFP

Scores an incoming RFP 1-5 on fit, budget signal, relationship, effort and incumbent risk from the facts you supply, and recommends pursue, pursue with conditions, or decline.

ChatGPTFees & contractsPractice management

Act as a skeptical practice principal. Below are an RFP summary and our firm
profile. Score this opportunity 1-5 on each: fit with our built experience,
realistic budget signal, relationship with the client (none / warm / existing),
submission effort for the stated deadline, and probability a local or
incumbent firm is favored. Justify each score in one sentence using only the
facts given. Do not invent facts about our firm or the client. End with a
one-line recommendation: pursue, pursue with conditions, or decline - and the
single strongest reason.

Firm profile: [3-5 lines: size, sectors, comparable projects, location]
RFP summary: [paste or summarize the RFP]
Tested with GPT-6 Astra,

From the guide: /business/chatgpt-architecture-proposals/

P–38ChatGPTPractice management

RFP compliance matrix for an architecture proposal

Extracts every requirement from a pasted RFP into a numbered matrix - exact wording quoted, mandatory vs scored, evidence to attach - with ambiguities listed as client questions, never guessed.

ChatGPTFees & contractsPractice management

Read the RFP below. Build a compliance matrix as a numbered table with these
columns: requirement (quote the RFP's exact wording), section of our response
that must answer it, mandatory or scored, and what evidence we need to attach.
Include every submission formality: deadline, delivery method, page limits,
required forms, insurance certificates, license numbers. Do not paraphrase
requirements - quote them. If anything is ambiguous or missing from the RFP,
list it under "Questions for the client" instead of guessing.

[paste the full RFP text]
Tested with GPT-6 Astra,

From the guide: /business/chatgpt-architecture-proposals/

P–39ChatGPTPractice management

RFP compliance matrix for a landscape commission

Builds a compliance matrix from a pasted RFP and firm facts - every requirement quoted, gaps flagged, agency questions listed - without inventing projects or staff.

ChatGPTFees & contractsPractice management

I'm a landscape architect responding to an RFP. Below I paste the RFP text
and bullet-point facts about my firm (size, relevant projects,
certifications, team). Build: (1) a compliance matrix - every stated
requirement, submission item and deadline from the RFP, each with the exact
RFP wording quoted and where my firm facts answer it or a gap flag; (2) a
response outline matching the RFP's required structure, with my facts
slotted in; (3) a list of questions worth sending to the issuing agency
before the deadline. Use only the RFP text and my facts - do not invent
projects, staff or certifications. RFP: [paste] Firm facts: [paste]
Tested with GPT-6 Astra,

From the guide: /guides/chatgpt-for-landscape-architects/

P–40ChatGPTPractice management

AI policy training quiz

Writes 8 scenario questions with answer keys tied to the rule letters of your office AI policy, for the staff training session.

ChatGPTPractice management

Our architecture office ([number] people) is adopting a written AI policy. Write 8 scenario-based training questions to check that staff understood it. The policy's key rules: (a) only approved tools for project data; (b) model-training opt-out on before first use, or a business plan where training is excluded by default; (c) never paste client-confidential documents, personal data or NDA material into unapproved tools; (d) AI output on codes, products, quantities and contract language is a draft verified by a qualified person against the source; (e) significant AI use on a project is logged; (f) no AI imitation of another architect's recognisable work; (g) report mistakes and near-misses without blame. Each question: a realistic office scenario, 3 answer options, the correct option, and a one-sentence explanation tied to the rule letter. Keep each question under 70 words.
Tested with Claude Sonnet 5 and GPT-6 Astra,

From the guide: /business/ai-policy-for-architecture-firms/

P–41ChatGPTPractice management

AI use log row from notes

Turns messy notes into one AI use log row for the project record, writing MISSING instead of inventing, and lists follow-up questions.

ChatGPTPractice management

Turn these messy notes into one row of an AI use log for a project record. Columns: Date | Project | Person | Tool and plan | Task | Data entered (and confidentiality level) | Output used where | Verification done (by whom, against what) | Risk tier. If a field is missing from the notes, write MISSING rather than inventing it. Then list the missing fields as questions to ask the person.

Notes: "[paste your notes]"
Tested with Claude Sonnet 5 and GPT-6 Astra,

From the guide: /business/ai-policy-for-architecture-firms/

P–42ChatGPTPractice management

Monday portfolio status digest

Turns scattered weekly notes into a Monday project digest ordered by attention needed, parking unplaceable notes under review by hand.

ChatGPTPractice management

Turn my scattered notes from last week into a Monday project digest. For
each project: status in one line, decisions waiting on me (and who is
waiting), deadlines in the next two weeks, and anything that reads urgent.
Order projects by how much attention they need today. Put notes you cannot
place under "review by hand" instead of guessing. Notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–43ChatGPTPractice management

Office chat-assistant ground rules memo

Drafts a one-page office memo of chat-assistant ground rules: what never gets pasted, account settings, the draft-verification rule and who answers questions.

ChatGPTPractice management

Draft a one-page memo of ground rules for using chat assistants in an
architecture office of [size]. Cover: material that never gets pasted
(client identifying data, anything under NDA, unpublished financials);
account settings to change before first use, naming the model-training
toggle; the rule that anything touching codes, products or contract language
is a draft a licensed professional verifies; and who in the office answers
questions. Plain language. I will align it with our insurer and counsel
before issuing.
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–44ChatGPTPractice management

Website project description in the studio voice

Writes a website project description from facts in the studio's own voice sample, without unsupported superlatives.

ChatGPTPractice management

Write a project description for my studio's website from the facts I paste:
type, size, location in general terms, design approach, key decisions,
materials. Match the voice of the sample description I include. 150 to 220
words. No superlatives the facts do not support, no client name unless I
marked it approved. Facts: [paste] Voice sample: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–45ChatGPTDesign development

Product data sheet comparison scaffold

Builds a product comparison table from two data sheets, flagging criteria tested under different standards instead of converting between them.

ChatGPTProduct researchDesign development

I paste two product data sheets for [product type], separated by ===. Build
a comparison table using only values that appear on the sheets: criterion,
product A, product B, and "not stated" where a sheet is silent. Criteria:
[list]. Flag any criterion where the two sheets report values under
different test standards, and do not convert between standards. Sheet A:
[paste] === Sheet B: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–46ChatGPTConcept

Area program vs lease limit check

Sums a room program, applies a grossing factor, tests it against a lease or zoning cap, and shows the arithmetic for every step.

ChatGPTSite analysisConcept

Here is my room program. 1) Sum the net areas. 2) Apply a grossing factor of [factor] to get gross floor area. 3) The limit is [limit] m2 gross - state whether we fit, and by how much we are over or under. 4) If we are over, calculate how much net area we must cut to fit. 5) State the net-to-gross efficiency as a percentage. Show the arithmetic step by step. I will paste the program below, all areas in m2 net.
Tested with GPT-6 Astra,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–47ChatGPTDesign development

Plant palette screen against site conditions

Screens a draft plant palette you wrote - name errors, zone fit, mature size vs space, invasiveness flags - without adding plants of its own.

ChatGPTSite analysisDesign development

Screen my draft plant palette, do not add plants of your own. I paste my
candidate list and the site conditions (location, USDA zone or local
equivalent, soil, sun, moisture, space available, maintenance level). For
each plant return a table row: botanical name as I wrote it; whether that
name looks current or possibly a synonym/misspelling (flag, don't rename);
stated zone fit vs my zone; mature size vs the space I listed; soil and sun
fit; one risk to verify (invasiveness in my region, disease, availability).
End with the list of flags I must verify against my regional flora,
extension service or nursery availability before this palette goes in a
drawing. Site: [paste] Palette: [paste]
Tested with GPT-6 Astra,

From the guide: /guides/chatgpt-for-landscape-architects/

P–48ChatGPTPre-design

Site visit notes into an observation report

Turns dictated site visit notes into an observation report, keeping every measurement as dictated and parking ambiguities under items to verify.

ChatGPTSite analysisPre-design

Turn my phone-dictated site visit notes into a site observation report under:
Access and approach / Topography and drainage / Existing structures and
condition / Utilities observed / Neighboring context / Vegetation / Items to
verify with a survey or authority. Keep every measurement exactly as I
dictated it; where my notes are ambiguous, list the item under Items to
verify instead of resolving it. Do not infer typical conditions for this
region. Notes: [paste]
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–49ChatGPTDocumentation

Outline spec section skeleton with VERIFY slots

Drafts a three-part outline specification skeleton as questions, with [VERIFY] slots where standards and test values belong.

ChatGPTSpecificationsDocumentation

Draft an outline specification skeleton for [section]. Structure only:
Part 1 General (submittals, quality assurance, warranties), Part 2 Products
(materials, components, finishes), Part 3 Execution (preparation,
installation, tolerances, protection) - as headings with the questions I
must answer under each. Write [VERIFY] wherever a standard, test value or
code reference belongs; do not supply the reference from memory. One page
maximum.
Tested with GPT-6 Astra,

From the guide: /prompts/chatgpt-prompts-for-architects/

P–50ChatGPTDocumentation

Plain-language client summary of a spec section

Translate a specification section into a 300-word plain-language summary for a client, with the binding-document disclaimer built in.

ChatGPTSpecificationsDocumentation

Translate the specification section below into a plain-language summary for a client who is not a construction professional. Explain what is being specified, the quality level chosen and what it means in practice, and which decisions are still open. Maximum 300 words. Add one line at the end stating that the specification document itself, not this summary, is the binding text.
Tested with GPT-6 Astra,

From the guide: /guides/ai-architectural-specifications/

P–51ChatGPTDocumentation

First-pass product research comparison table

Turn pasted manufacturer datasheet excerpts into a comparison table with exact stated values, gaps marked, and a [VERIFY] row - no outside product knowledge.

ChatGPTSpecificationsDocumentation

I will paste excerpts from manufacturer datasheets for [product type, e.g. mineral wool insulation boards] from several manufacturers. Build a comparison table: manufacturer, product line, the performance values stated in each excerpt (with the exact wording), and gaps where an excerpt does not state a value. Add a [VERIFY] row listing what must be confirmed on current manufacturer documentation before any of this enters a specification. Use only the pasted excerpts - no outside knowledge of these products.
Tested with GPT-6 Astra,

From the guide: /guides/ai-architectural-specifications/

P–52ClaudePre-design

Brief audit with conflict arithmetic

Full brief to program table, quantified requirements with their source sentences, internal conflicts with the arithmetic shown, and open client decisions.

ClaudeBrief analysisPre-design

I am an architect reviewing a client brief for a new project. I will paste the full brief below. After reading it, give me: 1) the stated space program with areas, as a table; 2) every quantitative requirement (areas, budgets, dates, parking) with the sentence it came from; 3) requirements that conflict with each other, with the arithmetic shown; 4) decisions the brief leaves open that the client must make before schematic design. Quote the brief verbatim for anything you flag, so I can check it in the source.
Tested with Claude Sonnet 5,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–53ClaudePre-design

Brief gap questions before SD

Lists the questions a client must answer before schematic design, grouped by budget, site, operations and quality, with what each blocks.

ClaudeBrief analysisPre-design

I will paste a project brief below. List the questions a client must answer before schematic design can start, grouped: budget and program; site and context; operations and users; quality and sustainability expectations. For each, say in one line what it blocks if unanswered. Only ask what the brief genuinely leaves open.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–54ClaudePre-design

Structured client-brief review

Turn a full client brief into a program table, quantified requirements, conflicts and open decisions — with verbatim quotes you can check.

ClaudeBrief analysisPre-design

I am an architect reviewing a client brief for a new project. I will paste the full brief below. After reading it, give me: 1) the stated space program with areas, as a table; 2) every quantitative requirement (areas, budgets, dates, performance targets) with the page or section it came from; 3) requirements that conflict with each other; 4) decisions the brief leaves open that the client will need to make before schematic design. Quote the brief verbatim for anything you flag, so I can check it in the source.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–55ClaudePre-design

RFP go / no-go summary

One-page summary of an RFP: scope, evaluation criteria, qualifications, deadlines and anything unusual to price before committing.

ClaudeBrief analysisPre-design

I am deciding whether my firm should respond to an RFP. I will paste the full RFP below. Summarize: 1) scope of services; 2) evaluation criteria with any stated weights; 3) minimum qualifications and required team roles; 4) submission requirements, page limits and deadlines; 5) anything unusual for this project type that I should price or clarify before committing. Keep it to one page and cite the RFP section for each point.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–56ClaudePre-design

RFP screen with contract-risk flags

One-page RFP summary sized to your firm: scope, weighted criteria, qualifications, deadlines, and liability terms to price or clarify before committing.

ClaudeBrief analysisPre-design

I am deciding whether my firm should respond to an RFP. I will paste the relevant excerpt below. Summarize on one page: 1) scope of services; 2) evaluation criteria with stated weights; 3) minimum qualifications; 4) submission requirements and deadlines; 5) anything unusual or risky for a firm my size that I should price or clarify before committing - contract liability terms included. Cite the RFP section for each point.
Tested with Claude Sonnet 5,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–57ClaudePre-design

Site analysis synthesis with cited observations

Turn your survey notes, verified zoning extract, climate notes and site visit observations into a cited observation table, opportunities and risks, and a missing-information checklist.

ClaudeBrief analysisPre-design

I am an architect starting site analysis for a [project type] on the site described
below. I will paste my raw material: survey notes, the zoning extract I have
verified, climate notes, and my own site visit observations.

Build a site analysis synthesis with three parts:
1. A table of observations: each site condition, its source in my material
   (quote the line), and the design question it raises.
2. Opportunities and risks: two plain lists, each item tied to a specific
   observation - no generic entries that could apply to any site.
3. Missing information: what an architect would still need to verify before
   schematic design, as a checklist.

Use only the material I paste. If something is not in my material, put it in
the missing-information list instead of guessing. Do not propose a design.

My material: [paste]
Tested with Claude Sonnet 5,

From the guide: /guides/ai-site-analysis/

P–58ClaudePre-design

Space program from meeting notes

Turns raw client-meeting notes into a space program table with [implied] and [NEEDS INPUT] flags - no invented areas.

ClaudeBrief analysisPre-design

Below are my raw notes from client meetings about a [project type]. Draft a space program table: space name, quantity, target area, key adjacencies, special requirements. Where the notes imply a space without naming it, include it flagged [implied]. Where area targets are missing, write [NEEDS INPUT] - do not invent numbers.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–59ClaudeAny stage

Client email about a slipped milestone

Draft a delay email that opens with the revised plan, explains the cause in two sentences, and protects the end date — under 250 words.

ClaudeClient communicationAny stage

Draft an email to my client. Situation: a design milestone will slip by [number] weeks because [reason]. Requirements: open with the revised plan rather than the problem; explain the cause in two sentences without blaming anyone; propose what we will progress in parallel so the end date is protected as far as possible; close with a specific next step and date. Professional, direct, no apologetic filler. Under 250 words. I will review and personalize before sending.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–60ClaudeAny stage

Raw notes to meeting minutes

Structured minutes with decisions, open questions and an action table; implied owners and deadlines marked [confirm] instead of guessed.

ClaudeClient communicationAny stage

Turn my raw meeting notes, pasted below, into minutes with: attendees; decisions made; open questions; and an action table with columns for item, owner, deadline. Where the notes imply an owner or deadline without stating one, mark it [confirm] rather than guessing. Keep the wording factual - no interpretation of anyone's intent.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–61ClaudeDesign development

Accessibility requirements screen

Turns pasted accessibility-standard excerpts into clause-numbered yes/no questions about your design - a structured screen, not a check.

ClaudeCode & zoningDesign development

I will paste excerpts from the accessibility standard applying to my project, plus a description of the spaces in question. Using only the pasted text, list each requirement with its clause number and phrase it as a yes/no question about my design. Mark questions the pasted text cannot settle "needs full standard". This structures my review - it does not replace it.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–62ClaudeDocumentation

Code Q&A restricted to pasted text

Ask code questions against excerpts you paste, with exact clause quotes — and 'not covered' instead of answers from general knowledge.

ClaudeCode & zoningDocumentation

I will paste excerpts from the building code that applies to my project, including the edition year and jurisdiction. Answer my questions using ONLY the text I provide. For every answer, quote the exact clause and its number. If the pasted text does not cover my question, say "not covered in the provided text" - do not answer from general knowledge. First question: list every requirement in these excerpts that applies to [describe the element], as a checklist with clause references.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–63ClaudePre-design

Code research brief, not a compliance check

Turns a project description into a research map: code chapters to open, project-specific questions, and a verification list with edition and renumbering risks - no settled numbers.

ClaudeCode & zoningPre-design

I am an architect starting code research for a project. Act as a research assistant, not a compliance checker. Project: [one-paragraph description - building use, storeys, construction type if known, jurisdiction and adopted code edition]. Produce: (1) a list of the code chapters and section families I most likely need to review for this project type, each with one line on why; (2) for each, the project-specific question I need to answer; (3) a verification list - what I must confirm in the adopted local code and amendments before relying on anything, including the edition year and renumbering risk between editions. Do not state final requirements as settled numbers; frame everything as sections to open and verify.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-building-code-checks/

P–64ClaudeDocumentation

Verification log for AI code answers

Converts AI code answers with section numbers into a source-check log table - claim, edition, section, what to verify, risk if wrong - without judging correctness.

ClaudeCode & zoningDocumentation

I asked an AI assistant some building-code questions and got answers with section numbers. Turn the answers I paste below into a verification log table with columns: claim, code edition cited, section cited, what to check at the source, risk if wrong, status (blank for me to fill). Add a final row for jurisdiction amendments. Do not tell me whether the claims are correct - the log is for my own source check.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-building-code-checks/

P–65ClaudeDocumentation

Local-amendment checklist for code sections

Builds a verify-everything checklist for specific code sections: where state and city amendments publish, per-section questions, renumbering risk between editions, who to contact.

ClaudeCode & zoningDocumentation

I am an architect preparing to rely on building-code research. Build me a local-amendment checklist for the following code sections: [code edition and section numbers], in a US jurisdiction whose state adopts the IBC with amendments. List, as a checklist: (1) where state and city amendments are typically published; (2) the specific questions to ask about each of these sections; (3) the renumbering risk - which of these section numbers changed between editions; (4) who to contact when the adopted edition is unclear. Frame every item as something I verify, not something you confirm.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-building-code-checks/

P–66ClaudePre-design

Zoning ordinance defined-terms extraction

Before any zoning analysis: lists every defined or ambiguous term in the pasted ordinance text and what to look up in the definitions chapter - catches missing height datums and undefined FAR.

ClaudeCode & zoningPre-design

Before I analyze a zoning ordinance, extract every defined or ambiguous term from the ordinance text I paste, and for each term tell me: where it is used, whether the pasted text defines it, and what I need to look up in the full ordinance's definitions chapter before relying on any analysis. Use ONLY the pasted text; do not supply definitions from general knowledge.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-building-code-checks/

P–67ClaudePre-design

Zoning first-pass screening table

Two-column table of pasted zoning requirements vs. your proposed values, flagged 'appears to meet / appears short / unclear' — a screen, not a compliance check.

ClaudeCode & zoningPre-design

I will paste the text of a zoning ordinance section and the basic parameters of a site and proposed building (use, lot area, frontage, height, setbacks, parking). Build a two-column table: each zoning requirement from the pasted text with its clause number, next to my proposed value, marked "appears to meet", "appears short" or "unclear from provided text". This is a first-pass screen to organize my own review. Do not state that anything complies.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–68ClaudeConstruction admin

Field report from raw site notes

Structures your site-visit notes into a factual field report - observations only, no directives, no acceptance of work.

ClaudeConstruction admin

Turn my raw site visit notes, pasted below, into a field report: project, date, weather, work observed in progress; items requiring attention with location and the drawing reference where I noted one; open questions for the contractor; photos referenced by my note numbers. Factual wording only - observations, no directives and no acceptance of work.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–69ClaudeConstruction admin

Punch list from walkthrough notes

Groups 140 phone notes into a numbered punch list by room and trade, keeping your wording and marking vague items [CLARIFY].

ClaudeConstruction admin

Organize my raw walkthrough notes, pasted below, into a punch list: items grouped by room, then by trade; each item as location, deficiency, reference to spec or drawing where my notes include one. Keep my wording for deficiencies - clean it only for clarity. Number items. Mark entries too vague to action [CLARIFY] so I can revisit them on site.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–70ClaudeConstruction admin

RFI response draft with citations

Drafts an RFI response that answers only what was asked, cites the controlling clause, and flags document conflicts - you review before issue.

ClaudeConstruction admin

I will paste a contractor RFI and the relevant drawing notes and specification clauses. Draft a response that: answers only what was asked; cites the controlling document and clause for each point; states "to be confirmed" where my pasted material does not settle the question. Neutral contractual tone. Note at the end if the question suggests a document conflict I should review. I will review and edit before issuing.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–71ClaudeConstruction admin

Submittal review against spec clauses

Pairs each pasted spec requirement with what the submittal shows - conform / deviation / not addressed - flagged vs unflagged deviations separated.

ClaudeConstruction admin

I will paste 1) the specification clauses governing a submittal and 2) the submitted product data. List: each spec requirement next to what the submittal shows, marked "appears to conform", "deviation" or "not addressed"; deviations the contractor flagged versus unflagged ones. Quote both documents. This organizes my review - the review action and stamp are mine.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–72ClaudeDesign development

Consultant coordination question list

Per-discipline lists of what your drawings must answer for each consultant and what you need from them, blockers first.

ClaudeCoordinationDesign development

I will describe the current state of a [project type] design. For each discipline - structural, mechanical, electrical, plumbing, [add others] - list the questions my drawings must answer for that consultant to progress, and the decisions I need FROM them before my next issue. Mark items that typically block others first. Base it only on what I describe.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–73ClaudeDesign development

Decision log from coordination notes

Converts meeting notes into a decision log plus open-items table; implied decisions stay [confirm], never promoted to decided.

ClaudeCoordinationDesign development

Turn my coordination meeting notes, pasted below, into a decision log: decision made, who made it, what it affects, drawings or documents to update; a second table of open items with owner and deadline. Where the notes imply a decision without stating it cleanly, put it under open items marked [confirm] - never promote it to decided.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–74ClaudeDesign development

Value-engineering impact notes

Per-item VE review scaffold: what each substitution changes, which design intent it touches, what to verify before accepting - [REFER] for consultant scope.

ClaudeCoordinationDesign development

I will paste a value-engineering proposal list. For each item, using only what I paste plus the design intent I describe: what the substitution changes in appearance, performance and maintenance; which stated design intent it touches; what I would need to verify before accepting (testing, warranty, code acceptance). Neutral wording I can send to a client after my review. Mark items needing consultant input [REFER].
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–75ClaudeConcept

Concept narrative from bullet points

Drafts a 400-600 word concept narrative from your facts and a style sample, marking every gap [NEEDS INPUT] instead of inventing.

ClaudeDesign narrativesConcept

Draft a concept narrative for a [project type]. Inputs below: 1) bullet-point facts about site, program and my design moves; 2) one narrative my firm wrote previously, as a style reference - match its tone, never copy its content. Use only facts from input 1; where a topic lacks a fact, write [NEEDS INPUT]. 400-600 words, for a client audience.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–76ClaudeConcept

Design options comparison matrix

Builds a client-ready comparison table of 2-3 schemes - program fit, site response, cost drivers, risks - and ends with the decision questions.

ClaudeDesign narrativesConcept

I will describe [number] design options for the same brief. Build a comparison table: how each option answers the program; site response; structural implications I listed; rough relative cost drivers; risks and unknowns. Use only what I describe - no invented benefits. Close with the questions the client must answer to choose, as a numbered list.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–77ClaudeConcept

Skeptical pre-review critique

A hostile-reviewer pass on your concept two days before the real review: intent vs execution gaps, circulation weaknesses, the three hardest questions.

ClaudeDesign reviewConcept

Act as a skeptical design reviewer. I will describe my concept for a [project type] below. Identify: where stated intent and described execution diverge; program requirements the scheme strains; circulation or organization weaknesses; claims a jury or planning officer would press hardest. Be specific and quote my own description back at me. End with the three hardest questions I should prepare for.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–78ClaudeAny stage

Rewrite in a practicing architect's voice

Strip marketing language and filler from a draft while keeping every fact, number and caveat exactly as given.

ClaudeDesign statementsAny stage

Rewrite the draft text I paste below so it reads like a practicing architect wrote it for a client presentation. Keep every fact, number, dimension and caveat exactly as given. Remove marketing language, repeated ideas and filler phrases. Shorter is better. Flag any sentence whose meaning you were unsure about rather than guessing.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–79ClaudePre-design

Design statement first draft

Draft a planning design statement from your site facts, the authority's checklist and one past statement as a style reference — gaps marked [NEEDS INPUT].

ClaudeDesign statementsPre-design

Draft a design statement for a planning submission. I will give you three inputs below: 1) bullet-point facts about the site, context and proposal; 2) the local authority's list of what the statement must address; 3) one design statement my firm wrote previously, as a style reference - match its tone and sentence rhythm, never copy its content. Only use facts from input 1. Where a required topic from input 2 has no matching fact, write the heading and the note [NEEDS INPUT] instead of inventing anything.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–80ClaudeDocumentation

CD set QA checklist

Builds a construction-documents review checklist by pass - cross-referencing, dimensions, coordination failure points - it cannot see your drawings.

ClaudeDocumentation QADocumentation

Build a construction documents QA checklist for a [project type]: cross-referencing (plans to schedules to details to specs); dimensional completeness; typical coordination failure points between architectural and [consultant disciplines] drawings; common omissions for this building type. Checklist format, grouped by review pass. This structures my review of the set - it cannot see my drawings and verifies nothing itself.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–81ClaudeDocumentation

Drawing notes vs spec contradiction table

Contradictions between drawing notes and a spec excerpt as a table - topic, what each document says, which to correct - plus dead references flagged.

ClaudeDocumentation QADocumentation

Check the drawing notes against the specification excerpt below and list every contradiction you find as a table: topic, what the drawings say, what the spec says, and which document I should probably correct. Also flag any reference that points to a document that does not exist in the list provided. Do not invent contradictions that are not there.
Tested with Claude Sonnet 5,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–82ClaudeDocumentation

General notes consistency check

Finds contradictions, duplicates and spec-material strays in your drawing-set general notes, quoted by note number - list only, you edit.

ClaudeDocumentation QADocumentation

I will paste the general notes from my drawing set. Find: notes that contradict each other; duplicates with different wording; notes referencing standards or documents inconsistently; notes that belong in the specifications instead. Quote each instance with its note number. Do not rewrite - list for my edit.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–83ClaudeDocumentation

pyRevit door-renumber script request

Asks Claude for a complete pyRevit script that renumbers doors by the room they open into - with the Transaction, fallback, reporting and DRY_RUN requirements spelled out so the output arrives reviewable.

ClaudeDocumentation QADocumentation

Write a complete pyRevit script (IronPython 2.7 compatible) that renumbers all doors in the
active Revit document by the room they open into: each door gets its ToRoom number (phase-aware,
last phase), with an A/B/C suffix when a room has several doors. Requirements: __title__ and
__doc__ header; imports from pyrevit (revit, DB, script); FilteredElementCollector over OST_Doors;
FromRoom fallback when ToRoom is empty; one named Transaction for all writes; write to the Mark
parameter with a read-only guard; skip and report doors with no room; include a DRY_RUN flag
that previews the result without writing. After the script, list your assumptions and what I
must check before running it on a live model.
Tested with Claude Fable 5.1 (output statically reviewed against pyRevit conventions; run from a command-line client, one run, not executed in Revit),

From the guide: /guides/claude-in-revit/

P–84ClaudeDocumentation

Fire-rating audit over Revit MCP read tools

After connecting a read-tier Revit MCP server, runs a whole-model door audit: discovers how the fire rating is actually stored, classifies missing vs empty vs placeholder, and hands back verifiable element selections.

ClaudeDocumentation QADocumentation

Using only the Revit MCP read tools, audit this model: which doors are missing a fire rating?
Before querying, check how the fire rating is actually stored: the built-in type parameter,
a shared instance parameter, or another name. Query the whole model, not the active view.
Classify each door: parameter absent, value empty, placeholder value ("-", "N/A", "TBD"),
or rated. Report counts per bucket, the exact parameter name you checked, and the query scope.
Then select the unrated doors in Revit so I can verify them myself. Do not modify anything.
After the summary, list the ways this audit could be wrong or incomplete in this model.
Tested with Claude Fable 5.1 (plan cross-checked against the official Autodesk tool list; run from a command-line client, one run, not executed in a live Revit session),

From the guide: /guides/claude-in-revit/

P–85ClaudePractice management

Additional services letter

A collaborative-but-firm letter at the third scope change: what was included, what changed, why it sits outside the base fee, the authorization path.

ClaudeFees & contractsPractice management

Draft a letter to a client about additional services. Facts below: original scope summary; the requested changes to date with dates; the contract clause covering additional services, quoted. Tone: collaborative and firm. Structure: what was included; what has changed; why the new work sits outside the base fee; the proposed path (estimate first, then authorization). Under 400 words. I will review against the signed contract before sending.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–86ClaudePractice management

Indemnification clause risk review for counsel

Maps the risks in a client-drafted indemnification clause in plain language and builds the question list for your lawyer and insurance broker - never legal advice.

ClaudeFees & contractsPractice management

Review this indemnification clause from a client-drafted agreement for architectural services and tell me: 1) each specific risk it creates for my firm, in plain language; 2) how it differs from a negligence-based standard; 3) the questions I should take to my lawyer and insurance broker before signing. Do not rewrite the clause; do not present your analysis as legal advice - frame it as issues to raise with counsel. I will paste the clause below.
Tested with Claude Sonnet 5,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–87ClaudePractice management

AI-use risk-tier classifier

Sorts a list of office AI uses into Low, Medium, High or Prohibited, with the required check for each, and names the items it is unsure about.

ClaudePractice management

You are helping a small architecture practice classify its AI uses by risk. Use three tiers plus a prohibited category:
- Low: internal, no client data, quick reasonableness check by the user.
- Medium: client-facing or informs design decisions; output checked against a trusted source; noted in the project record; PM approval.
- High: informs stamped/sealed documents, life safety, accessibility, code interpretation, structural assumptions or contract language; AI only as an exploratory/checking aid; fully re-verified by a qualified person; logged; architect-of-record approval.
- Prohibited: anything barred by client contract or law, or pasting confidential client data into a tool not on the approved list.

Classify each of these uses. For each, give: tier, one-line reason, the required check. Output a markdown table only, then one line naming any item you were unsure about and why.

[paste your list of AI uses, one per line]
Tested with Claude Sonnet 5 and GPT-6 Astra,

From the guide: /business/ai-policy-for-architecture-firms/

P–88ClaudePractice management

AI vendor data-terms reader

Reads the data-use section of an AI tool's terms: training default, opt-out, retention, output ownership, sharing, business plan - quoted, NOT STATED where silent, with a verdict.

ClaudePractice management

Below is the data-use section of an AI tool's terms that a staff member wants the office to adopt. Read it and fill this table: | Question | Answer from the text | Exact quote | Risk for an architecture firm |. Questions: (1) Does the vendor train on our inputs by default? (2) Can we opt out, and how? (3) How long are inputs retained? (4) Who owns outputs? (5) Can the vendor share data with third parties? (6) Is there a business/commercial plan with different terms? If the text does not answer a question, write NOT STATED - do not guess. After the table, give a verdict in at most 3 lines: approve / approve with conditions / reject, and the conditions.

TERMS EXCERPT:
[paste the data-use section of the tool's terms]
Tested with Claude Sonnet 5 and GPT-6 Astra,

From the guide: /business/ai-policy-for-architecture-firms/

P–89ClaudePractice management

Staff AI-use one-pager

Drafts an office guideline for AI chat tools: what never gets pasted, required settings incl. training opt-out, and the draft-only rule for codes and contracts.

ClaudePractice management

Draft a one-page internal guideline for AI chat tools in an architecture practice of [size] in [country]. Cover: material that must never be pasted (client-confidential documents, personal data, anything under NDA); required account settings before use, with model-training opt-out named explicitly; the rule that AI output on codes, products and contracts is a draft requiring professional verification; who in the office answers questions. Plain language, fits one page. I will adapt it to our insurance and legal advice.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–90ClaudeDesign development

Product comparison from datasheets

Tables pasted manufacturer data for shortlisted products and forces a [VERIFY] row - availability, lead time, code acceptance, price are always calls.

ClaudeProduct researchDesign development

I will paste manufacturer information for [number] candidate products for [application]. Build a comparison table using only the pasted data: key performance values, dimensions, finish options, stated warranty, anything one datasheet covers that another omits. Add a [VERIFY] row listing claims I must confirm directly with each manufacturer - current availability, lead time, local code acceptance and price are always on it.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–91ClaudePre-design

Unfamiliar building-type research list

A pre-concept research checklist for a building type you have not done - regulatory bodies, adjacencies, standards, consultants - with [VERIFY LOCALLY] flags.

ClaudeResearchPre-design

I am an architect starting my first [building type] project in [country or region]. List what I should research before concept design: regulatory frameworks and review bodies typical for this type; typical functional requirements and adjacencies; common dimensional standards; typical consultants needed; the questions experienced designers of this type ask first. Format as a checklist I can work through, with [VERIFY LOCALLY] on anything jurisdiction-dependent.
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–92ClaudePre-design

Site constraints digest

Merges survey notes, deed and utility records into one constraints summary with per-item source quotes and a missing-information list.

ClaudeSite analysisPre-design

I will paste site information below (survey notes, deed description, utility records, my own observations). Build a site constraints summary: dimensions and area; topography; access; utilities; easements or restrictions; anything in the documents that limits buildable area. Quote the source document for each item and list what information is still missing for a [project type].
Tested with Claude Sonnet 5.5,

From the guide: /prompts/claude-prompts-for-architects/

P–93ClaudeDocumentation

Outline spec section in CSI 3-part format

An outline specification for one section (example: 08 11 13 hollow metal) with standards flagged [VERIFY EDITION] and no invented product numbers.

ClaudeSpecificationsDocumentation

Draft an outline specification for Section [number and title], in CSI 3-part format (Part 1 General, Part 2 Products, Part 3 Execution), for a [project type and location]. [Scope notes: interior/exterior, finishes, fire ratings.] Keep it an outline: numbered articles and short clauses, not full prose. Where you reference a standard (ANSI, NAAMM, UL and similar), add [VERIFY EDITION] after it, and do not invent product model numbers.
Tested with Claude Sonnet 5,

From the guide: /tools/claude-vs-chatgpt-for-architects/

P–94ClaudeDocumentation

Outline spec sections from a product schedule

Turn a product and materials schedule into MasterFormat outline sections with a [VERIFY] list — no invented products, standards or values.

ClaudeSpecificationsDocumentation

I will paste a product and materials schedule for a project. Draft outline specification sections organized by CSI MasterFormat division. For each section include: section number and title, a short description, the products from my schedule that belong there, and a [VERIFY] list of what must be confirmed with manufacturers or consultants before the section is developed. Do not invent product names, standards or performance values that are not in my schedule.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–95ClaudeDocumentation

Consistency check: drawing notes vs spec section

Find contradictions between drawing notes and a draft spec section - terminology, materials, dimensions, dead cross-references - as a findings table.

ClaudeSpecificationsDocumentation

Below are (A) the general and keyed notes from our drawings and (B) a draft specification section for the same scope. List every inconsistency between them: terminology that differs for the same item, materials mentioned in one but not the other, dimensions or performance requirements that conflict, and notes that reference spec sections that do not exist in the package list I provide. Format as a table: item / what A says / what B says / suggested resolution. Do not rewrite either document.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-architectural-specifications/

P–96ClaudeDocumentation

Edit an office master spec down to project scope

Cut an office master section down to one project's scope with every deletion and modification listed - no silent drops, no added requirements.

ClaudeSpecificationsDocumentation

I will paste a section from our office master specification, then a one-paragraph description of this project's scope. Edit the master down to what this project needs. List every paragraph you deleted and why, every paragraph you modified with before/after text, and anything in the master that looks outdated and should be checked. Do not add new requirements, products or standards.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-architectural-specifications/

P–97ClaudeDocumentation

Reference spec package gap review

Audit a previous project's spec package against a new scope: divisions covered, divisions missing, unusual or contradictory clauses quoted.

ClaudeSpecificationsDocumentation

I will paste a specification package from a previous project that we are using as a reference. For a project of this scope - [describe project type and scope in one sentence] - list: 1) which MasterFormat divisions are covered; 2) divisions this project will likely need that are missing; 3) sections whose requirements look unusual or contradictory, with the clause quoted. Review only - do not rewrite anything.
Tested with Claude Sonnet 5.5,

From the guide: /guides/claude-for-architects/

P–98ClaudeDocumentation

Pre-issue red-flag pass on a spec section

A skeptical senior-reviewer pass before issue: every product number, code citation, performance value, odd warranty and design-delegation clause flagged with the quote.

ClaudeSpecificationsDocumentation

Review the specification section below as a skeptical senior spec writer. Flag: any product name or model number (each needs manufacturer verification), any code or standard citation (each needs edition and jurisdiction verification), any performance value (each needs a source), warranty terms that differ from standard practice, and any clause that assigns design responsibility to the contractor. Return a numbered list with the quoted text and why it needs attention. You are a reviewer - change nothing.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-architectural-specifications/

P–99ClaudeDocumentation

Draft one spec section in CSI 3-part format

Draft a single specification section (General / Products / Execution) from project input only, with [VERIFY] placeholders instead of invented standards or values.

ClaudeSpecificationsDocumentation

Draft specification section [number and title, e.g. 08 11 13 Hollow Metal Doors and Frames] in CSI 3-part format (Part 1 General, Part 2 Products, Part 3 Execution) using ONLY the project information I paste below. Where standard practice requires a reference standard, a product, or a performance value that my input does not contain, write [VERIFY: what is needed] instead of a value. Do not cite any code section or ASTM/ANSI standard number unless it appears in my input.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-architectural-specifications/

P–100ClaudeDocumentation

Submittal register first pass from spec sections

Extract every submittal requirement from pasted spec sections into a register with clause quotes; flag missing submittal articles and duplicates.

ClaudeSpecificationsDocumentation

From the specification sections I paste below, extract every submittal requirement into a register: section number, submittal item, type (product data / shop drawings / samples / test reports / closeout), and the exact clause text that requires it. Flag any section that has requirements in Part 2 or Part 3 but no corresponding submittal article, and any duplicate requirements across sections. Extraction only - do not add requirements.
Tested with Claude Sonnet 5,

From the guide: /guides/ai-architectural-specifications/

P–101GeminiPre-design

Brief audit with conflicts and open decisions

A client brief reviewed for its space program, every quantitative requirement with its source sentence, internal conflicts with the arithmetic shown, and the decisions the client still owes you.

GeminiBriefs & programmingPre-design

I am an architect reviewing a client brief for a new project. I will paste
the full brief below. After reading it, give me: 1) the stated space program
with areas, as a table; 2) every quantitative requirement (areas, budgets,
dates, parking) with the sentence it came from; 3) requirements that
conflict with each other, with the arithmetic shown; 4) decisions the brief
leaves open that the client must make before schematic design. Quote the
brief verbatim for anything you flag, so I can check it in the source.
Tested with Gemini 4 Argon,

From the guide: /tools/chatgpt-vs-gemini-for-architects/

P–102GeminiPre-design

RFP go/no-go summary priced to your firm size

An RFP excerpt condensed to scope, weighted evaluation criteria, minimum qualifications, deadlines, and the contract or insurance terms a firm your size must price or question before committing.

GeminiBusiness developmentPre-design

I am deciding whether my [N]-person architecture firm should respond to an
RFP. I will paste the relevant excerpt below. Summarize on one page: 1) scope
of services; 2) evaluation criteria with stated weights; 3) minimum
qualifications; 4) submission requirements and deadlines; 5) anything unusual
or risky for a firm my size that I should price or clarify before committing.
Cite the RFP section for each point.
Tested with Gemini 4 Argon,

From the guide: /tools/chatgpt-vs-gemini-for-architects/

P–103GeminiConcept

Context aerial

You need to see a proposed building in its wider context from above.

GeminiClient communicationConcept

A 16:9 oblique aerial photograph, taken as if from a drone at about 60 metres, of a proposed [primary school / health centre] on a [riverside / edge-of-town] site. The building is a low two-storey courtyard scheme with a sawtooth roof over the larger wing, surrounded by existing [terraced housing / farmland / industrial sheds]. Materials: pale grey fibre-cement panels on the walls, a planted sedum roof on the flat sections, and timber-lined window reveals. Morning light from the east, slight haze towards the horizon. No people, no cars, no text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–104GeminiDocumentation

Exploded axonometric with labels

You need an exploded axonometric that explains a building's layers.

GeminiClient communicationDocumentation

A clean exploded axonometric diagram of a [three-storey timber-frame community building], viewed from the south-east, with the roof, each floor plate, the structural frame and the ground floor slab separated vertically and aligned on a common axis. Each layer labelled in a small sans-serif font with these exact labels: "Sedum roof", "CLT roof panels", "Second floor", "First floor", "Glulam frame", "Ground floor slab". Thin leader lines from each label to its layer. Flat pastel colours, white background, no shadows, no people, square 1:1 frame.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–105GeminiDesign development

Plan to isometric

You want a 2D floor plan turned into an isometric 3D view for a client who does not read plans.

GeminiClient communicationDesign development

Use the uploaded 2D floor plan as the exact layout. Produce a 3D isometric cutaway view of this plan seen from above at a 45-degree angle, with walls extruded to a consistent height of about one storey, all door and window openings in the positions shown, and no rooms added, removed, merged or resized. Keep room proportions exactly as drawn. Floors in pale oak, walls in matte white, simple furniture placed only where the plan indicates it. Clean white background, soft shadows, no text or dimensions, square 1:1 frame.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–106GeminiConcept

Proposal on the empty site photo

You want your proposed building dropped into a photo of the empty site.

GeminiClient communicationConcept

Use the uploaded site photograph as the base. Keep the sky, the neighbouring buildings, the road, the trees, the people and the lighting exactly as photographed; the only change is to add a new building on the [empty plot / fenced site] in the centre of the frame. The new building: a three-storey volume with a flat roof, a brick ground floor in pale buff brick, two upper floors in dark grey standing-seam metal, and regular vertical windows with deep reveals. Match the camera height, perspective, lens and light direction of the photograph. No text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–107GeminiDesign development

Cladding A/B

You want a cladding A/B on the same image.

GeminiDesign optionsDesign development

Use the uploaded image as the base. Keep the exact same massing, window positions, floor count, roofline, camera, lighting, site and sky. The only change: replace the [current cladding, e.g. white render] on the upper floors with [new cladding, e.g. vertical untreated larch boards weathered to silver-grey], with the same board direction across every elevation visible in the image. Leave the ground floor, the window frames and everything else exactly as shown. No text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–108GeminiConcept

Three massing options, one image

You want three massing alternatives for the same site in a single image.

GeminiDesign optionsConcept

A single 16:9 image divided into three equal panels, side by side, each showing the same [rectangular / L-shaped] site from the same high aerial viewpoint at the same scale. Left panel: one long linear block along the north edge with a courtyard to the south. Middle panel: two parallel blocks with a shared garden between them. Right panel: a compact square block in the centre of the site with open space on all four sides. All three volumes are three storeys, rendered as plain white model massing with no windows, on a neutral grey site plate with existing neighbouring buildings shown in light grey. Soft studio lighting, no text or labels.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–109GeminiConcept

First exterior concept

You have a site and a programme and want a first exterior image to react to.

GeminiSite analysisConcept

A wide 16:9 eye-level photograph of a two-storey [house / community library / small office] on a [sloping suburban / flat rural / tight urban infill] site in [city or region]. The building is a simple rectangular volume with a single-pitch roof falling towards the street, a recessed entrance on the left third of the facade, and a full-height glazed bay on the right. Materials: pale lime-washed brick walls, a dark standing-seam zinc roof, and untreated larch boards on the recessed entrance. Mature trees behind the building, a gravel forecourt in front. Late afternoon sun from the left, long soft shadows. No people, no cars, no text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–110GeminiDocumentation

Outline spec with [VERIFY EDITION] tags

A CSI-format outline specification for one section, every referenced standard tagged [VERIFY EDITION], no invented product numbers. Gemini 4 Argon returned the widest reference list in our 12-task test.

GeminiSpecificationsDocumentation

Draft an outline specification for Section [number - title], in CSI 3-part
format (Part 1 General, Part 2 Products, Part 3 Execution), for a [project
type and location]. [Key conditions: interior/exterior, finishes, fire
ratings.] Keep it an outline: numbered articles and short clauses, not full
prose. Where you reference a standard (ANSI, NAAMM, UL and similar), add
[VERIFY EDITION] after it, and do not invent product model numbers.
Tested with Gemini 4 Argon,

From the guide: /tools/chatgpt-vs-gemini-for-architects/

P–111GeminiDesign development

Day to night

You need the same view at night for a lighting or planning conversation.

GeminiVisualizationDesign development

Use the uploaded image as the base. Keep the exact same building, massing, window positions, floor count, roofline, materials, site and camera. The only change is time of day: convert the scene to a clear evening about thirty minutes after sunset, deep blue sky, warm interior light visible through the windows on all floors, soft downlighting at the entrance, street lamps on, no illuminated signage. Keep all surroundings in their photographed positions. No people, no cars, no text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–112GeminiDesign development

Facade re-clad, openings kept

You need to show an existing facade re-clad without changing its openings.

GeminiVisualizationDesign development

Use the uploaded facade photograph as the base. Keep the exact same massing, window positions, window sizes, floor count, roofline, doors, camera and lighting from the photograph. The only change: replace the existing [painted render / tired brick / grey panel] cladding with [new material, e.g. glazed white terracotta baguettes] across the whole facade, with [slim bronze] window frames replacing the existing frames in the same openings. Leave the street, sky, neighbouring buildings and all people and vehicles exactly as photographed. No text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–113GeminiConcept

Model photo onto the real site

You have a photo of a physical study model and want to see it as a built building on the real site.

GeminiVisualizationConcept

Image A is a photograph of a physical study model; Image B is a photograph of the real site. Use Image A for the building's massing, proportions, window openings and roofline, kept exactly as modelled; do not add floors or openings. Use Image B for the site, surroundings, horizon and lighting. Place the building on the site at the position of the [empty plot / existing car park] in Image B, matched to the camera height and perspective of Image B. Materials: [material 1] walls, [material 2] roof, [material 3] window frames. Keep everything else in Image B exactly as photographed. No people, no cars, no text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–114GeminiDesign development

Model screenshot, materials added

You have a SketchUp, Revit or Rhino screenshot and want materials and light added without geometry changes.

GeminiVisualizationDesign development

Use the uploaded model screenshot as the source of all geometry. Keep the exact same massing, window positions, floor count, roofline, column positions and camera from the screenshot; the only changes are materials, lighting and context. Apply these materials: [material 1, e.g. board-formed concrete] on the ground floor, [material 2, e.g. vertical charred-timber cladding] on the upper floors, [material 3, e.g. slim black aluminium] window frames. Replace the blank model ground with a paved forecourt and low planting. Golden hour backlighting creating long shadows, slight lens flare. No people, no cars, no text, no watermark. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/

P–115GeminiConcept

Sketch to photoreal view

You have a hand-drawn exterior perspective and want a photoreal version of exactly that view.

GeminiVisualizationConcept

Use the uploaded hand sketch as the only source of geometry. Keep the exact same massing, window positions, floor count, roofline, proportions and camera angle from the sketch. Do not add floors, change the roof pitch, or move openings. Render the building as a finished photograph: walls in rough textured render painted warm white, the roof in dark grey clay tiles, window frames in dark timber. Site as drawn, with the existing [tree / wall / neighbouring building] kept in place. Soft late-morning light from the right. No people, no cars, no text, no watermark. 3:2 landscape frame. Photorealistic architectural photography.
Tested with Checked against Google's Nano Banana Pro prompt guidance (20 Nov 2025) and current docs (20 Oct 2026); not run through the image models,

From the guide: /prompts/gemini-nano-banana-prompts-for-architects/