Benchmark Almanac

SCADBench

Gemini 3.1 Pro 1,685 leads since February 2026
Not in model cards

Models write OpenSCAD code from the same short prompt. Voters rotate two rendered objects and choose which one better fits the prompt.

The maker ranks 19 models by Elo. The gallery lists 46 prompts, including a ruler, an anchor, and a watch.

Made by
SCADBench
Kind
CAD models
Level
University
Scored
Ranked by votes
Answers
Code
Tasks
All public
Measures
Preference
Language
English
In and out
Text → Code
Interaction
Single turn
Items from
Live play or votes
Refresh
Live
Run by
A host
In views
CAD, Kept fresh
Strongest model tested
Claude Opus 4.6 #39 on the capability index
Last model added
February 2026
Scores refreshed
November 20th, 2026
19
Models scored on its own leaderboard

Gemini 3.1 Pro leads with 1,685 among models released by February 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

The newest model with a score here came out in February 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

Bars start at 800.

  1. 1 Gemini 3.1 Pro 1,685
  2. 2 Claude Opus 4.6 1,440
  3. 3 Gemini 3 Flash 1,399
  4. 4 GPT-5.2 1,353
  5. 5 GLM-5 1,341
  6. 6 Claude Sonnet 4.6 1,339
  7. 7 Kimi K2.5 1,329
  8. 8 Gemini 2.5 Pro Jun 2025 1,277
  9. 9 Gemini 2.5 Flash Sep 2025 1,179
  10. 10 Claude Haiku 4.5 1,166
  11. 11 Grok 4.1 Fast 1,152
  12. 12 MiniMax-M2.5 1,120
  13. 13 Qwen3.5-Plus 1,087
  14. 14 GPT-4o Nov 2024 1,069
  15. 15 DeepSeek-V3.2 1,069
  16. 16 Llama 4 Maverick 994
  17. 17 Gemini 2.5 Flash-Lite Sep 2025 945
  18. 18 Llama 4 Scout 932
  19. 19 Nemotron 3 Nano 30B-A3B 917

Source: Official SCADBench leaderboard

#ModelScoreSettingDateSource
1 Gemini 3.1 Pro 1,685 Feb 2026 Official leaderboard
2 Claude Opus 4.6 1,440 Feb 2026 Official leaderboard
3 Gemini 3 Flash 1,399 Dec 2025 Official leaderboard
4 GPT-5.2 1,353 Dec 2025 Official leaderboard
5 GLM-5 1,341 Feb 2026 Official leaderboard
6 Claude Sonnet 4.6 1,339 Feb 2026 Official leaderboard
7 Kimi K2.5 1,329 Jan 2026 Official leaderboard
8 Gemini 2.5 Pro Jun 2025 1,277 Jun 2025 Official leaderboard
9 Gemini 2.5 Flash Sep 2025 1,179 Sep 2025 Official leaderboard
10 Claude Haiku 4.5 1,166 Oct 2025 Official leaderboard
11 Grok 4.1 Fast 1,152 Nov 2025 Official leaderboard
12 MiniMax-M2.5 1,120 Feb 2026 Official leaderboard
13 Qwen3.5-Plus 1,087 Feb 2026 Official leaderboard
14 GPT-4o Nov 2024 1,069 Nov 2024 Official leaderboard
15 DeepSeek-V3.2 1,069 Dec 2025 Official leaderboard
16 Llama 4 Maverick 994 Apr 2025 Official leaderboard
17 Gemini 2.5 Flash-Lite Sep 2025 945 Sep 2025 Official leaderboard
18 Llama 4 Scout 932 Apr 2025 Official leaderboard
19 Nemotron 3 Nano 30B-A3B 917 Dec 2025 Official leaderboard

The best score, 1,685, costs $0.15 a task

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Meta Google DeepMind DeepSeek Anthropic Cost frontier One model at each effort level
0 500 1,000 1,500 2,000 $0.0001 $0.001 $0.01 $0.10 $1.00 Cost per task, log scale Gemini 3.1 Pro Claude Opus 4.6 Gemini 3 Flash Gemini 2.5 Flash DeepSeek-V3.2 Gemini 2.5 Flash-Lite 0 500 1,000 1,500 2,000 $0.0001 $0.001 $0.01 $0.10 $1.00 Cost per task, log scale Gemini 3.1 Pro Claude Opus 4.6

Source: Official SCADBench leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 3.1 Pro 1,685 $0.15
Claude Opus 4.6 1,440 $0.03
Gemini 3 Flash 1,399 $0.002
GPT-5.2 1,353 $0.04
GLM-5 1,341 $0.008
Claude Sonnet 4.6 1,339 $0.02
Kimi K2.5 1,329 $0.01
Gemini 2.5 Pro Jun 2025 1,277 $0.07
Gemini 2.5 Flash Sep 2025 1,179 $0.001
Claude Haiku 4.5 1,166 $0.004
Grok 4.1 Fast 1,152 $0.002
MiniMax-M2.5 1,120 $0.002
Qwen3.5-Plus 1,087 $0.002
GPT-4o Nov 2024 1,069 $0.004
DeepSeek-V3.2 1,069 $0.0004
Llama 4 Maverick 994 $0.0003
Gemini 2.5 Flash-Lite Sep 2025 945 $0.0003
Llama 4 Scout 932 $0.0002
Nemotron 3 Nano 30B-A3B 917 $0.0008

6 models have held number one on the official leaderboard since 2023

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI Google DeepMind Anthropic
Own leaderboard Not tested since Feb 2026

Source: Official SCADBench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
800 1,000 1,200 1,400 1,600 1,800 2025 2026 800 1,000 1,200 1,400 1,600 1,800 2025 2026

Source: Official SCADBench leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Engineering and CAD

All 112 in Engineering and CAD
  • EEBench In use 13 electrical-engineering circuit and PCB design tasks graded by deterministic build checks, simulation and BOM constraints. No LLM judge. Gemini 4 Carbon 83.9%
  • 3DCodeBench In use Agentic procedural 3D modeling via code: author engine-ready 3D assets through software APIs. Graded for executability and shape fidelity. LongCat-2.5 55.4%
  • CADGenBench Only in system cards Agents construct CAD models from design prompts. The generation split is graded for executability and geometric correctness. Claude Fable 5 45.1%
  • PartBench Only in system cards Parametric mechanical part design and CAD-style generation, graded for geometric and manufacturing fidelity. Grok 4.6 58.5%
  • Razavi-Bench Not in model cards Models explain what happens in analog circuits and why. A language model awards up to four points per answer against a reference and a rubric. Gemini 4 Carbon 98.3%
  • CircuitRubric Not in model cards Models write analog circuit netlists from design requests. A program compares transistor connections and relative device sizes with accepted reference circuits. Claude Opus 4.8 96%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 9 months ago, in February 2026, for Gemini 3.1 Pro. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (19 entries).
  • Pass Models scored 19 models we track.
  • Warning Newest score The newest came 9 months ago, in February 2026, for Gemini 3.1 Pro.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 4.6 (#39).
  • Note Headroom Scored as an Elo rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on scadbench.com: 19 rows for models we track at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check