Benchmark Almanac

PCBWorld-Bench

PPO (w/o finish ) 0.79 tops its leaderboard GPT-5.4 0.54 leads the models we track
Not in model cards

Agents route printed circuit boards in KiCad. A board earns a clean pass when all required connections are complete and KiCad reports zero design-rule errors.

The authors report the fraction of boards earning a clean pass. The benchmark includes held-out synthetic layouts and 679 open hardware boards.

Released
July 2026
Kind
Electronics
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Actions
Interaction
Agentic
Environment
KiCad
Refresh
One fixed set
Run by
Anyone
In views
Agents
Strongest model tested
GPT-5.4 #28 on the capability index
Last model added
March 2026
Scores refreshed
November 20th, 2026
8
Entries in its paper, 4 models we track

PPO (w/o finish ) tops the leaderboard with 0.79 GPT-5.4 leads with 0.54 among models released by March 2026

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The paper reports no overall score, so this is the plain mean of the clean-pass fractions on D2, D3-A and D3-B. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GPT-5.4, at 0.54. This view shows only models we track, each at its best.

The newest model with a score here came out in March 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 PPO (w/o finish ) Best of five rollouts 0.79
  2. 2 PPO Best of five rollouts 0.77
  3. 3 PPO (terminal) Best of five rollouts 0.71
  4. 4 GRPO Best of five rollouts 0.65
  5. 5 GPT-5.4 Best of five rollouts 0.54
  6. 6 GPT-5.4 Mini Best of five rollouts 0.29
  7. 7 GPT-5.4 Nano Best of five rollouts 0.28
  8. 8 Qwen3.5 397B-A17B Best of five rollouts 0.22

Source: The PCBWorld-Bench paper

#EntryScoreSettingDateSource
1 PPO (w/o finish ) 0.79 Best of five rollouts Official leaderboard
2 PPO 0.77 Best of five rollouts Official leaderboard
3 PPO (terminal) 0.71 Best of five rollouts Official leaderboard
4 GRPO 0.65 Best of five rollouts Official leaderboard
5 GPT-5.4 0.54 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
6 GPT-5.4 Mini 0.29 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
7 GPT-5.4 Nano 0.28 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
8 Qwen3.5 397B-A17B 0.22 Best of five rollouts, PCBWorld LLM agent Feb 2026 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.4 0.54 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
2 GPT-5.4 Mini 0.29 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
3 GPT-5.4 Nano 0.28 Best of five rollouts, PCBWorld LLM agent Mar 2026 Official leaderboard
4 Qwen3.5 397B-A17B 0.22 Best of five rollouts, PCBWorld LLM agent Feb 2026 Official leaderboard

Rankings by part

We read 3 parts from the benchmark’s paper: D2, D3-A and D3-B. Each ranks models on one part of the test. GPT-5.4 leads 2 of them, and Qwen3.5 397B-A17B leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

D2

Fraction of boards with a clean pass. 4 models we track have a score here.

  1. 1 GPT-5.4 0.96
  2. 2 GPT-5.4 Mini 0.58
  3. 3 GPT-5.4 Nano 0.55
  4. 4 Qwen3.5 397B-A17B 0.33

Number one over time

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

OpenAI
Benchmark paper Not tested since Jul 2026
GPT-5.4 0.54

Source: The PCBWorld-Bench paper

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
In the benchmark’s paper
0.20 0.30 0.40 0.50 0.60 2026 0.20 0.30 0.40 0.50 0.60 2026

Source: Official PCBWorld-Bench 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 8 months ago, in March 2026, for GPT-5.4 Mini. 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 benchmark’s paper (8 entries).
  • Pass Models scored 4 models we track, plus 4 other entries on its leaderboard.
  • Warning Newest score The newest came 8 months ago, in March 2026, for GPT-5.4 Mini.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.4 (#28).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 4 rows for models we track and 4 other entries 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