Benchmark Almanac

OmniLayout

Gemini 2.5 Flash-Lite Sep 2025 0.76 leads since June 2026
Not in model cards

Models place components inside a board outline. The evaluator checks overlap, routing, and electrical functionality against the schematic and the reference layout.

Nine general-purpose models appear in the placement tables. The authors withdrew all 1,681 placement graphs on October 26, 2026 for a license review.

Released
June 2026
Kind
Electronics
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text, Images → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
A host
In views
Vision
Strongest model tested
GPT-5.5 #21 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
9
Models scored on its own leaderboard

Gemini 2.5 Flash-Lite leads with 0.76 among models released by April 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 April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Gemini 2.5 Flash-Lite Sep 2025, All Tools- 0.76
  2. 2 Gemini 3.1 Pro Visualization 0.71
  3. 3 GPT-5.5 Sch. SW-Net. 0.59
  4. 4 Claude Sonnet 4.6 Sch. SW-Net. 0.58
  5. 5 Qwen3.5-9B Sch. SW-Net. 0.58
  6. 6 Qwen 3.6 35B-A3B Sch. SW-Net. 0.57
  7. 7 Llama 4 Maverick All Tools 0.55
  8. 8 GPT-5 mini Sch. SW-Net. 0.53
  9. 9 Ministral 3 14B Sch. SW-Net. 0.50

Source: Official OmniLayout leaderboard

#ModelScoreSettingDateSource
1 Gemini 2.5 Flash-Lite Sep 2025 0.76 All Tools- Sep 2025 Official leaderboard
2 Gemini 3.1 Pro 0.71 Visualization Feb 2026 Official leaderboard
3 GPT-5.5 0.59 Sch. SW-Net. Apr 2026 Official leaderboard
4 Claude Sonnet 4.6 0.58 Sch. SW-Net. Feb 2026 Official leaderboard
5 Qwen3.5-9B 0.58 Sch. SW-Net. Feb 2026 Official leaderboard
6 Qwen 3.6 35B-A3B 0.57 Sch. SW-Net. Apr 2026 Official leaderboard
7 Llama 4 Maverick 0.55 All Tools Apr 2025 Official leaderboard
8 GPT-5 mini 0.53 Sch. SW-Net. Aug 2025 Official leaderboard
9 Ministral 3 14B 0.50 Sch. SW-Net. Dec 2025 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Gemini 2.5 Flash-Lite
Visualizati… Fewshot Sch. Img All Tools-
0.76
Gemini 3.1 Pro
Zeroshot Fewshot Visualizati…
0.71
GPT-5.5
Zeroshot Sch. Img
0.59
Claude Sonnet 4.6
All Tools Zeroshot Sch. Img Sch. SW-Net.
0.58
Qwen3.5-9B
Zeroshot Fewshot Sch. Sem. Sch. SW-Net.
0.58
Qwen 3.6 35B-A3B
Sch. Img Zeroshot Sch. SW-Net.
0.57
Llama 4 Maverick
Geo. Check Fewshot Sch. SW-Net. Visualizati…
0.55
GPT-5 mini
All Tools Fewshot Zeroshot Sch. SW-Net.
0.53
Ministral 3 14B
All Tools Geo. Check Visualizati…
0.50

Source: Official OmniLayout leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 2.5 Flash-Lite Sep 2025 All Tools- 0.76
Gemini 3.1 Pro Visualization 0.71
GPT-5.5 Sch. SW-Net. 0.59
GPT-5.5 Visualization 0.59
Claude Sonnet 4.6 Sch. SW-Net. 0.58
GPT-5.5 Sch. Sem. 0.58
Gemini 3.1 Pro Sch. Img 0.58
Qwen3.5-9B Sch. SW-Net. 0.58
Qwen 3.6 35B-A3B Sch. SW-Net. 0.57
GPT-5.5 All Tools 0.57
GPT-5.5 Sch. Img 0.56
Claude Sonnet 4.6 Visualization 0.55
Gemini 3.1 Pro Geo. Check 0.55
Gemini 3.1 Pro Sch. SW-Net. 0.55
Gemini 3.1 Pro Sch. Sem. 0.55
GPT-5.5 Routing Score 0.55
Gemini 3.1 Pro Fewshot 0.55
Claude Sonnet 4.6 Sch. Sem. 0.55
Llama 4 Maverick All Tools 0.55
Claude Sonnet 4.6 Sch. Img 0.54
GPT-5.5 Geo. Check 0.54
Qwen 3.6 35B-A3B Fewshot 0.54
GPT-5.5 Fewshot 0.54
GPT-5 mini Sch. SW-Net. 0.53
Llama 4 Maverick Visualization 0.53
Gemini 3.1 Pro Routing Score 0.53
Qwen 3.6 35B-A3B Sch. Sem. 0.52
Qwen3.5-9B Sch. Sem. 0.52
GPT-5.5 Zeroshot 0.52
Gemini 3.1 Pro All Tools 0.52
Claude Sonnet 4.6 Routing Score 0.52
Qwen 3.6 35B-A3B Zeroshot 0.52
Claude Sonnet 4.6 Geo. Check 0.52
GPT-5 mini Sch. Sem. 0.51
Llama 4 Maverick Sch. Sem. 0.51
Llama 4 Maverick Sch. Img 0.51
Gemini 2.5 Flash-Lite Sep 2025 Geo. Check 0.51
Ministral 3 14B Sch. SW-Net. 0.50
Gemini 2.5 Flash-Lite Sep 2025 Sch. Img 0.50
Gemini 3.1 Pro Zeroshot 0.50
GPT-5 mini Visualization 0.50
Ministral 3 14B Sch. Sem. 0.49
GPT-5 mini Sch. Img 0.49
Ministral 3 14B Zeroshot 0.49
Llama 4 Maverick Sch. SW-Net. 0.49
Claude Sonnet 4.6 Fewshot 0.49
GPT-5 mini Zeroshot 0.49
Gemini 2.5 Flash-Lite Sep 2025 Zeroshot 0.49
Ministral 3 14B Sch. Img 0.49
Claude Sonnet 4.6 Zeroshot 0.49
Gemini 2.5 Flash-Lite Sep 2025 Sch. SW-Net. 0.48
Llama 4 Maverick Zeroshot 0.48
Qwen 3.6 35B-A3B Sch. Img 0.48
Ministral 3 14B Fewshot 0.47
Ministral 3 14B Visualization 0.47
GPT-5 mini Geo. Check 0.47
Gemini 2.5 Flash-Lite Sep 2025 Sch. Sem. 0.46
Gemini 2.5 Flash-Lite Sep 2025 Fewshot 0.45
Llama 4 Maverick Fewshot 0.45
Qwen3.5-9B Fewshot 0.45
GPT-5 mini Fewshot 0.44
GPT-5 mini Routing Score 0.43
Ministral 3 14B Routing Score 0.42
Ministral 3 14B Geo. Check 0.42
Claude Sonnet 4.6 All Tools 0.41
Gemini 2.5 Flash-Lite Sep 2025 Routing Score 0.41
Gemini 2.5 Flash-Lite Sep 2025 Visualization- 0.40
Qwen3.5-9B Zeroshot 0.40
GPT-5 mini All Tools 0.40
Ministral 3 14B All Tools 0.38
Llama 4 Maverick Routing Score 0.37
Llama 4 Maverick Geo. Check 0.36

Number one over time

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

Google DeepMind
Own leaderboard Not tested since Jun 2026

Source: Official OmniLayout 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
0.40 0.50 0.60 0.70 0.80 2025 2026 0.40 0.50 0.60 0.70 0.80 2025 2026

Source: Official OmniLayout 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 7 months ago, in April 2026, for GPT-5.5. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (9 entries).
  • Pass Models scored 9 models we track.
  • Warning Newest score The newest came 7 months ago, in April 2026, for GPT-5.5.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.5 (#21).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on omnieda.com: 72 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