How well can AI read construction drawings?

We ask the models 169 questions across 9 real drawing sets, like "What's the concrete spec for the F2 footings?" and "How many F2 footings are there?", and count how many they get right.

169

Questions

Real drawing sets

Raw models

Harness + model

82.1%

Best model alone

93.3%

Best in a harness

Inside a harness

#	Harness	Model	Accuracy	vs raw	Details
1	Cowork	Claude Sonnet 5.5	93.3%	+32.8	›
Cowork · Claude Sonnet 5.5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 60.5%. Reading the sheet 96.5% 29 questions Schedules 100.0% 13 questions Counting 91.2% 34 questions Element detail 95.2% 21 questions Scale and measurement 81.8% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 100.0% 7 questions Knowing when to stop 33.3% 3 questions
2=	Cowork	Claude Fable 5.1	92.5%	+22.4	›
Cowork · Claude Fable 5.1 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 70.2%. Reading the sheet 100.0% 29 questions Schedules 92.3% 13 questions Counting 91.2% 34 questions Element detail 95.2% 21 questions Scale and measurement 72.7% 11 questions Cross-referencing 92.3% 13 questions Scope 100.0% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 33.3% 3 questions
3=	Cowork	Claude Opus 5.5	92.5%	+21.6	›
Cowork · Claude Opus 5.5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 70.9%. Reading the sheet 96.5% 29 questions Schedules 100.0% 13 questions Counting 85.3% 34 questions Element detail 100.0% 21 questions Scale and measurement 72.7% 11 questions Cross-referencing 92.3% 13 questions Scope 100.0% 6 questions Sheet structure 100.0% 7 questions Knowing when to stop 100.0% 3 questions
4	Cowork	Claude Opus 5	91.8%	+23.9	›
Cowork · Claude Opus 5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 67.9%. Reading the sheet 100.0% 29 questions Schedules 100.0% 13 questions Counting 88.2% 34 questions Element detail 95.2% 21 questions Scale and measurement 81.8% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 33.3% 3 questions
5	ChatGPT Work	GPT 5.6 Sol	90.3%	+11.2	›
ChatGPT Work · GPT 5.6 Sol by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 79.1%. Reading the sheet 96.5% 29 questions Schedules 100.0% 13 questions Counting 85.3% 34 questions Element detail 100.0% 21 questions Scale and measurement 90.9% 11 questions Cross-referencing 76.9% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 33.3% 3 questions
6	Claude Code	Claude Opus 5	88.1%	+20.2	›
Claude Code · Claude Opus 5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 67.9%. Reading the sheet 93.1% 29 questions Schedules 100.0% 13 questions Counting 82.3% 34 questions Element detail 90.5% 21 questions Scale and measurement 63.6% 11 questions Cross-referencing 84.6% 13 questions Scope 100.0% 6 questions Sheet structure 100.0% 7 questions Knowing when to stop 0.0% 3 questions
7	ChatGPT Work	GPT 6 Astra	86.6%	+4.5	›
ChatGPT Work · GPT 6 Astra by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 82.1%. Reading the sheet 86.2% 29 questions Schedules 100.0% 13 questions Counting 79.4% 34 questions Element detail 95.2% 21 questions Scale and measurement 63.6% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 100.0% 7 questions Knowing when to stop 33.3% 3 questions
8=	ChatGPT Work	GPT 6.1 Sol	85.8%	—	›
ChatGPT Work · GPT 6.1 Sol by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. Reading the sheet 89.7% 29 questions Schedules 76.9% 13 questions Counting 79.4% 34 questions Element detail 90.5% 21 questions Scale and measurement 81.8% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 100.0% 7 questions Knowing when to stop 33.3% 3 questions
9=	Muse	Muse Spark 1.3	85.8%	—	›
Muse · Muse Spark 1.3 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. Reading the sheet 89.7% 29 questions Schedules 100.0% 13 questions Counting 82.3% 34 questions Element detail 85.7% 21 questions Scale and measurement 72.7% 11 questions Cross-referencing 84.6% 13 questions Scope 83.3% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 33.3% 3 questions
10	Gemini app	Gemini 4 Argon	85.1%	+8.9	›
Gemini app · Gemini 4 Argon by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 76.1%. Reading the sheet 96.5% 29 questions Schedules 69.2% 13 questions Counting 76.5% 34 questions Element detail 90.5% 21 questions Scale and measurement 81.8% 11 questions Cross-referencing 84.6% 13 questions Scope 100.0% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 33.3% 3 questions
11	Cursor	Kimi K3	83.6%	—	›
Cursor · Kimi K3 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. Reading the sheet 96.5% 29 questions Schedules 100.0% 13 questions Counting 73.5% 34 questions Element detail 85.7% 21 questions Scale and measurement 54.5% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 33.3% 3 questions
12	Cowork	Claude Sonnet 5	82.8%	+22.4	›
Cowork · Claude Sonnet 5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 60.5%. Reading the sheet 96.5% 29 questions Schedules 100.0% 13 questions Counting 70.6% 34 questions Element detail 90.5% 21 questions Scale and measurement 45.5% 11 questions Cross-referencing 84.6% 13 questions Scope 83.3% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 100.0% 3 questions
13=	Cursor	Claude Opus 5	82.1%	+14.2	›
Cursor · Claude Opus 5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 67.9%. Reading the sheet 96.5% 29 questions Schedules 84.6% 13 questions Counting 76.5% 34 questions Element detail 81.0% 21 questions Scale and measurement 54.5% 11 questions Cross-referencing 92.3% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 33.3% 3 questions
14=	Cursor	GLM 5.2	82.1%	—	›
Cursor · GLM 5.2 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. Reading the sheet 96.5% 29 questions Schedules 92.3% 13 questions Counting 79.4% 34 questions Element detail 90.5% 21 questions Scale and measurement 45.5% 11 questions Cross-referencing 69.2% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 33.3% 3 questions
15	ChatGPT Work	GPT 6 Sol	81.2%	—	›
ChatGPT Work · GPT 6 Sol by category All 169 questions in one session. The score covers the 133 questions the drawings can answer. 168 of 169 responses passed validation; accuracy covers 133 of 134 answerable questions. Original output preserved. Reading the sheet 89.7% 29 questions Schedules 92.3% 13 questions Counting 70.6% 34 questions Element detail 95.0% 20 questions Scale and measurement 45.5% 11 questions Cross-referencing 84.6% 13 questions Scope 100.0% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 100.0% 3 questions
16	Codex	GPT 5.6 Sol	80.6%	+1.5	›
Codex · GPT 5.6 Sol by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 79.1%. Reading the sheet 96.5% 29 questions Schedules 92.3% 13 questions Counting 67.7% 34 questions Element detail 95.2% 21 questions Scale and measurement 54.5% 11 questions Cross-referencing 61.5% 13 questions Scope 83.3% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 0.0% 3 questions
17=	Cursor	Grok 4.6	79.8%	+30.6	›
Cursor · Grok 4.6 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 49.3%. Reading the sheet 93.1% 29 questions Schedules 100.0% 13 questions Counting 70.6% 34 questions Element detail 81.0% 21 questions Scale and measurement 63.6% 11 questions Cross-referencing 69.2% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 0.0% 3 questions
18=	Cursor	GPT 5.6 Sol	79.8%	+0.8	›
Cursor · GPT 5.6 Sol by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 79.1%. Reading the sheet 93.1% 29 questions Schedules 76.9% 13 questions Counting 70.6% 34 questions Element detail 85.7% 21 questions Scale and measurement 63.6% 11 questions Cross-referencing 76.9% 13 questions Scope 83.3% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 33.3% 3 questions
19	Cowork	Claude Haiku 5.5	78.4%	+15.7	›
Cowork · Claude Haiku 5.5 by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 62.7%. Reading the sheet 93.1% 29 questions Schedules 84.6% 13 questions Counting 58.8% 34 questions Element detail 95.2% 21 questions Scale and measurement 54.5% 11 questions Cross-referencing 69.2% 13 questions Scope 100.0% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 100.0% 3 questions
20	Cursor	Gemini 3.7 Flash	76.9%	+6.7	›
Cursor · Gemini 3.7 Flash by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 70.2%. Reading the sheet 93.1% 29 questions Schedules 100.0% 13 questions Counting 55.9% 34 questions Element detail 90.5% 21 questions Scale and measurement 45.5% 11 questions Cross-referencing 61.5% 13 questions Scope 100.0% 6 questions Sheet structure 85.7% 7 questions Knowing when to stop 33.3% 3 questions
21	ChatGPT Work	GPT 5.6 Luna	69.4%	+1.5	›
ChatGPT Work · GPT 5.6 Luna by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 67.9%. Reading the sheet 96.5% 29 questions Schedules 69.2% 13 questions Counting 47.1% 34 questions Element detail 90.5% 21 questions Scale and measurement 45.5% 11 questions Cross-referencing 46.2% 13 questions Scope 83.3% 6 questions Sheet structure 71.4% 7 questions Knowing when to stop 0.0% 3 questions
22	ChatGPT Work	GPT 6 Luna	27.6%	−44.0	›
ChatGPT Work · GPT 6 Luna by category All 169 questions in one session. The score covers the 134 questions the drawings can answer. The same model through the API scored 71.6%. Reading the sheet 55.2% 29 questions Schedules 30.8% 13 questions Counting 8.8% 34 questions Element detail 42.9% 21 questions Scale and measurement 0.0% 11 questions Cross-referencing 7.7% 13 questions Scope 50.0% 6 questions Sheet structure 14.3% 7 questions Knowing when to stop 100.0% 3 questions

Each pairing ran once, and repeat runs move by 1.1 to 5.4 points on their own.

The models on their own

#	Model	Lab	Accuracy	Cost	Details
1	GPT 6 Astra openai/gpt-6-astra	OpenAI	82.1%	$24.11	›
GPT 6 Astra by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 6.7% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 73.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 95.2% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 45.5% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 84.6% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
2	GPT 5.6 Sol openai/gpt-5.6-sol	OpenAI	79.1%	$13.54	›
GPT 5.6 Sol by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 1.5% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 64.7% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 36.4% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 76.9% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
3	Muse Wonder meta/muse-wonder	Meta	77.6%	$10.09	›
Muse Wonder by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 11.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 61.8% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 36.4% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
4=	Gemini 4 Argon google/gemini-4-argon	Google	76.1%	$2.06	›
Gemini 4 Argon by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 16.4% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 61.8% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 27.3% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 33.3% 3 questions Questions the drawings don't answer, where the right answer is to say so.
4=	GPT 5.6 Terra openai/gpt-5.6-terra	OpenAI	76.1%	$5.56	›
GPT 5.6 Terra by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 3.7% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 70.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 84.6% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
6	GPT 6 Sol openai/gpt-6-sol	OpenAI	75.4%	$5.18	›
GPT 6 Sol by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 6.7% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 64.7% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 27.3% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 84.6% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 71.4% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
7=	Gemini 3.8 Flash google/gemini-3.8-flash	Google	71.6%	$1.18	›
Gemini 3.8 Flash by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 96.5% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 47.1% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 33.3% 3 questions Questions the drawings don't answer, where the right answer is to say so.
7=	GPT 6 Luna openai/gpt-6-luna	OpenAI	71.6%	$0.31	›
GPT 6 Luna by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 6.7% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 55.9% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 84.6% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
9	Claude Opus 5.5 anthropic/claude-opus-5.5	Anthropic	70.9%	$5.69	›
Claude Opus 5.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 13.4% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 84.6% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 47.1% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 27.3% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
10=	Claude Fable 5.1 anthropic/claude-fable-5.1	Anthropic	70.2%	$15.05	›
Claude Fable 5.1 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 35.3% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 36.4% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 84.6% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 66.7% 3 questions Questions the drawings don't answer, where the right answer is to say so.
10=	Gemini 3.7 Flash google/gemini-3.7-flash	Google	70.2%	$0.33	›
Gemini 3.7 Flash by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 11.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 100.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 41.2% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 53.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
12	GPT 5.5 openai/gpt-5.5	OpenAI	69.4%	$19.90	›
GPT 5.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 6.0% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 42.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 27.3% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 76.9% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 35.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
13=	Claude Opus 5 anthropic/claude-opus-5	Anthropic	67.9%	$7.45	›
Claude Opus 5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 20.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 35.3% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 36.4% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 61.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
13=	GPT 5.6 Luna openai/gpt-5.6-luna	OpenAI	67.9%	$0.41	›
GPT 5.6 Luna by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 5.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 50.0% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 42.9% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 66.7% 3 questions Questions the drawings don't answer, where the right answer is to say so.
15=	Claude Fable 5 anthropic/claude-fable-5	Anthropic	65.7%	$14.71	›
Claude Fable 5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 13.4% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 93.1% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 84.6% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 32.4% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 53.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 66.7% 3 questions Questions the drawings don't answer, where the right answer is to say so.
15=	Claude Opus 4.8 anthropic/claude-opus-4.8	Anthropic	65.7%	$7.06	›
Claude Opus 4.8 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 27.6% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 26.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 90.5% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 27.3% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
17	Gemini 3.6 Flash google/gemini-3.6-flash	Google	63.4%	$3.38	›
Gemini 3.6 Flash by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.5% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 84.5% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 27.9% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 83.3% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 13.6% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 69.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
18=	Claude Haiku 5.5 anthropic/claude-haiku-5.5	Anthropic	62.7%	$0.16	›
Claude Haiku 5.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 23.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 32.4% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 76.2% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 61.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
18=	Gemini 3.1 Pro Preview google/gemini-3.1-pro-preview	Google	62.7%	$2.46	›
Gemini 3.1 Pro Preview by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 18.7% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 92.3% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 29.4% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 61.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 71.4% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
20=	Claude Sonnet 5 anthropic/claude-sonnet-5	Anthropic	60.5%	$2.94	›
Claude Sonnet 5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 26.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 89.7% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 69.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 26.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 46.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
20=	Claude Sonnet 5.5 anthropic/claude-sonnet-5.5	Anthropic	60.5%	$2.66	›
Claude Sonnet 5.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 31.3% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 35.3% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 76.2% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 46.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 50.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 100.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
22	Claude Opus 4.7 anthropic/claude-opus-4.7	Anthropic	57.5%	$6.48	›
Claude Opus 4.7 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 23.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 20.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 85.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 61.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 66.7% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
23=	Gemini 3.1 Flash Lite google/gemini-3.1-flash-lite	Google	56.7%	$0.08	›
Gemini 3.1 Flash Lite by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 82.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 69.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 26.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 76.2% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 61.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 42.9% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
23=	GPT 5.4 Mini openai/gpt-5.4-mini	OpenAI	56.7%	$2.28	›
GPT 5.4 Mini by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 11.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 75.9% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 35.3% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 71.4% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 53.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
25=	Claude Sonnet 4.5 anthropic/claude-sonnet-4.5	Anthropic	51.5%	$4.14	›
Claude Sonnet 4.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 28.4% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 53.8% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 20.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 71.4% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 38.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 66.7% 3 questions Questions the drawings don't answer, where the right answer is to say so.
25=	GPT 4.1 openai/gpt-4.1	OpenAI	51.5%	$1.61	›
GPT 4.1 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 11.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 86.2% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 53.8% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 23.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 81.0% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 38.5% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 100.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 14.3% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
27=	Gemini 2.5 Pro google/gemini-2.5-pro	Google	49.3%	$2.29	›
Gemini 2.5 Pro by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 79.3% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 76.9% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 23.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 57.1% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 30.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 66.7% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
27=	Grok 4.6 x-ai/grok-4.6	xAI	49.3%	$10.71	›
Grok 4.6 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 24.6% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 65.5% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 61.5% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 32.4% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 38.1% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 18.2% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 53.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 85.7% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 33.3% 3 questions Questions the drawings don't answer, where the right answer is to say so.
29=	Claude Haiku 4.5 anthropic/claude-haiku-4.5	Anthropic	46.3%	$1.57	›
Claude Haiku 4.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 34.3% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 82.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 46.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 14.7% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 76.2% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 30.8% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 28.6% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
29=	Gemini 2.5 Flash google/gemini-2.5-flash	Google	46.3%	$0.36	›
Gemini 2.5 Flash by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 14.9% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 82.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 61.5% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 20.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 52.4% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 46.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 50.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 28.6% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
31	Qwen3 VL 235b A22b Instruct qwen/qwen3-vl-235b-a22b-instruct	Alibaba	44.8%	$1.18	›
Qwen3 VL 235b A22b Instruct by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 36.6% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 82.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 46.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 23.5% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 66.7% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 15.4% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 14.3% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
32	GPT 5 Nano openai/gpt-5-nano	OpenAI	41.8%	$0.19	›
GPT 5 Nano by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 38.8% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 75.9% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 46.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 17.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 57.1% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 23.1% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 83.3% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 14.3% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
33	Grok 4.5 x-ai/grok-4.5	xAI	40.3%	$2.12	›
Grok 4.5 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 25.4% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 58.6% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 61.5% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 14.7% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 42.9% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 9.1% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 46.2% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 66.7% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 57.1% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
34	GPT 4o openai/gpt-4o	OpenAI	38.8%	$1.79	›
GPT 4o by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 33.6% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 72.4% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 38.5% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 11.8% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 71.4% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 15.4% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 66.7% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 14.3% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
35	Gemini 2.5 Flash Lite google/gemini-2.5-flash-lite	Google	37.3%	$0.08	›
Gemini 2.5 Flash Lite by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 26.1% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 65.5% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 46.2% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 17.6% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 47.6% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 23.1% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 50.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 42.9% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 66.7% 3 questions Questions the drawings don't answer, where the right answer is to say so.
36=	Mistral Large 2512 mistralai/mistral-large-2512	Mistral	9.0%	$0.13	›
Mistral Large 2512 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 67.2% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 13.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 0.0% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 8.8% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 9.5% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 15.4% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 16.7% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 0.0% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.
36=	Mistral Medium 3.1 mistralai/mistral-medium-3.1	Mistral	9.0%	$0.12	›
Mistral Medium 3.1 by category Answered all 169 questions. The score covers the 134 the drawings can answer. It declined 73.5% of those. On site, a model that won't commit is as much of a problem as one that guesses. Reading the sheet 13.8% 29 questions Finding a fact the drawing states outright, in the title block, the general notes or the legend. Schedules 7.7% 13 questions Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet. Counting 8.8% 34 questions Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill. Element detail 9.5% 21 questions The size, type or material of a named element, like a beam size, a pipe material or a cable size. Scale and measurement 0.0% 11 questions Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere. Cross-referencing 7.7% 13 questions Answers that need more than one sheet, or tracing a single line diagram through. Scope 0.0% 6 questions What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one. Sheet structure 14.3% 7 questions Reading the sheet itself: how many details or grid lines it has, and what revision it's at. Knowing when to stop 100.0% 3 questions Questions the drawings don't answer, where the right answer is to say so.

Cost is one run of all 169 questions.

The categories

What we asked them to do

The questions split into 9 kinds of work, and some are much harder than others. A model can read a schedule well and still not be able to count.

Reading the sheet

32 questions

Finding a fact the drawing states outright, in the title block, the general notes or the legend.

Schedules

18 questions

Pulling one value out of a table such as a door schedule, panel schedule or equipment list. Easy to get wrong when the table runs across half a sheet.

Counting

41 questions

Counting tagged items on a plan, like fittings or outlets. This is the core takeoff skill.

Element detail

25 questions

The size, type or material of a named element, like a beam size, a pipe material or a cable size.

Scale and measurement

22 questions

Distances, areas and volumes the model has to measure off the drawing at scale, because the number isn't printed anywhere.

Cross-referencing

14 questions

Answers that need more than one sheet, or tracing a single line diagram through.

Scope

7 questions

What's in and what's out: new or existing, who supplies it, and whether it's in this package or another one.

Sheet structure

7 questions

Reading the sheet itself: how many details or grid lines it has, and what revision it's at.

Knowing when to stop

3 questions

Questions the drawings don't answer, where the right answer is to say so.

We re-run this on every major release

Every new model goes through the same 169 questions. Join the community to get the results as they come in, and the workflows we build from them.

Get the results first →

How well can AI read construction drawings?

Inside a harness

Cowork · Claude Sonnet 5.5 by category

Cowork · Claude Fable 5.1 by category

Cowork · Claude Opus 5.5 by category

Cowork · Claude Opus 5 by category

ChatGPT Work · GPT 5.6 Sol by category

Claude Code · Claude Opus 5 by category

ChatGPT Work · GPT 6 Astra by category

ChatGPT Work · GPT 6.1 Sol by category

Muse · Muse Spark 1.3 by category

Gemini app · Gemini 4 Argon by category

Cursor · Kimi K3 by category

Cowork · Claude Sonnet 5 by category

Cursor · Claude Opus 5 by category

Cursor · GLM 5.2 by category

ChatGPT Work · GPT 6 Sol by category

Codex · GPT 5.6 Sol by category

Cursor · Grok 4.6 by category

Cursor · GPT 5.6 Sol by category

Cowork · Claude Haiku 5.5 by category

Cursor · Gemini 3.7 Flash by category

ChatGPT Work · GPT 5.6 Luna by category

ChatGPT Work · GPT 6 Luna by category

The models on their own

GPT 6 Astra by category

GPT 5.6 Sol by category

Muse Wonder by category

Gemini 4 Argon by category

GPT 5.6 Terra by category

GPT 6 Sol by category

Gemini 3.8 Flash by category

GPT 6 Luna by category

Claude Opus 5.5 by category

Claude Fable 5.1 by category

Gemini 3.7 Flash by category

GPT 5.5 by category

Claude Opus 5 by category

GPT 5.6 Luna by category

Claude Fable 5 by category

Claude Opus 4.8 by category

Gemini 3.6 Flash by category

Claude Haiku 5.5 by category

Gemini 3.1 Pro Preview by category

Claude Sonnet 5 by category

Claude Sonnet 5.5 by category

Claude Opus 4.7 by category

Gemini 3.1 Flash Lite by category

GPT 5.4 Mini by category

Claude Sonnet 4.5 by category

GPT 4.1 by category

Gemini 2.5 Pro by category

Grok 4.6 by category

Claude Haiku 4.5 by category

Gemini 2.5 Flash by category

Qwen3 VL 235b A22b Instruct by category

GPT 5 Nano by category

Grok 4.5 by category

GPT 4o by category

Gemini 2.5 Flash Lite by category

Mistral Large 2512 by category

Mistral Medium 3.1 by category