Benchmark Almanac

Closed-Loop Evaluation of LLM Agents for Embedded Software Development

GPT-5.4 100% leads since March 2026
Not in model cards

Agents write firmware from engineering briefs and test it on simulated equipment. Hidden checks score device behavior across five control tasks.

The authors publish 420 runs for seven models under four feedback conditions. In the primary condition, agents test their own work before one graded submission.

Kind
Electronics
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Code → Code, Actions
Interaction
Agentic
Environment
Terminal, A simulator
Items from
Written by experts
Refresh
One fixed set
Run by
Anyone
In views
Agents, Robotics
Strongest model tested
GPT-5.4 #28 on the capability index
Last model added
March 2026
Scores refreshed
November 20th, 2026
7
Models scored on its own leaderboard

GPT-5.4 leads with 100% among models released by March 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 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 GPT-5.4 ci_red_green 100%
  2. 2 Qwen3.5-27B oracle_full 80%
  3. 3 GPT-5.4 Mini ci_red_green 80%
  4. 4 Qwen3.5-35B-A3B oracle_full 66.7%
  5. 5 Qwen3.5-9B oracle_full 46.7%
  6. 6 Qwen3.5-4B oracle_full 13.3%
  7. 7 Qwen3.5-2B ci_red_green 0%

Source: Official Closed-Loop Evaluation of LLM Agents for Embedded Software Development leaderboard

#ModelScoreSettingDateSource
1 GPT-5.4 100% ci_red_green Mar 2026 Official leaderboard
2 Qwen3.5-27B 80% oracle_full Feb 2026 Official leaderboard
3 GPT-5.4 Mini 80% ci_red_green Mar 2026 Official leaderboard
4 Qwen3.5-35B-A3B 66.7% oracle_full Feb 2026 Official leaderboard
5 Qwen3.5-9B 46.7% oracle_full Feb 2026 Official leaderboard
6 Qwen3.5-4B 13.3% oracle_full Feb 2026 Official leaderboard
7 Qwen3.5-2B 0% ci_red_green Feb 2026 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
GPT-5.4
oneshot_bli… ci_red_green
100%
Qwen3.5-27B
oneshot_bli… oracle_full
80%
GPT-5.4 Mini
realistic_s… oneshot_bli… ci_red_green
80%
Qwen3.5-35B-A3B
oneshot_bli… ci_red_green oracle_full
66.7%
Qwen3.5-9B
oneshot_bli… oracle_full
46.7%
Qwen3.5-4B
oneshot_bli… oracle_full
13.3%
Qwen3.5-2B
ci_red_green
0%

Source: Official Closed-Loop Evaluation of LLM Agents for Embedded Software Development leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-5.4 ci_red_green 100%
GPT-5.4 oracle_full 93.3%
GPT-5.4 realistic_self_verify 93.3%
GPT-5.4 oneshot_blind 86.7%
Qwen3.5-27B oracle_full 80%
GPT-5.4 Mini ci_red_green 80%
GPT-5.4 Mini oracle_full 73.3%
Qwen3.5-35B-A3B oracle_full 66.7%
GPT-5.4 Mini oneshot_blind 66.7%
Qwen3.5-27B ci_red_green 60%
Qwen3.5-27B realistic_self_verify 60%
Qwen3.5-27B oneshot_blind 53.3%
Qwen3.5-35B-A3B ci_red_green 53.3%
GPT-5.4 Mini realistic_self_verify 53.3%
Qwen3.5-9B oracle_full 46.7%
Qwen3.5-35B-A3B oneshot_blind 26.7%
Qwen3.5-35B-A3B realistic_self_verify 26.7%
Qwen3.5-9B ci_red_green 20%
Qwen3.5-4B oracle_full 13.3%
Qwen3.5-4B realistic_self_verify 13.3%
Qwen3.5-9B oneshot_blind 13.3%
Qwen3.5-9B realistic_self_verify 13.3%
Qwen3.5-4B ci_red_green 6.70%
Qwen3.5-2B ci_red_green 0%
Qwen3.5-2B oneshot_blind 0%
Qwen3.5-2B oracle_full 0%
Qwen3.5-2B realistic_self_verify 0%
Qwen3.5-4B oneshot_blind 0%

2 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.

Qwen OpenAI
Own leaderboard Passed 85% in Mar 2026, by GPT-5.4
GPT-5.4 100%

Source: Official Closed-Loop Evaluation of LLM Agents for Embedded Software Development leaderboard

GPT-5.4 on the official leaderboard passed 85% in March 2026

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 Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.4 Qwen3.5-27B 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.4 Qwen3.5-27B

Source: Official Closed-Loop Evaluation of LLM Agents for Embedded Software Development 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

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the official leaderboard (7 entries).
  • Pass Models scored 7 models we track.
  • Note Newest score The newest came 8 months ago, in March 2026, for GPT-5.4 Mini. That’s expected for a saturated test.
  • Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is GPT-5.4 (#28).
  • Note Headroom Saturated. The best is 100%, past 85% since March 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 28 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