Benchmark Almanac

RTL-Repo

Qwen3 236B A22B 41.9% tops its leaderboard Qwen3-Coder-480B-A35B 38.9% leads the models we track
Not in model cards

Models predict the next line of Verilog using code from the same project. The authors measure exact matches and edit similarity.

TuRTLe publishes exact-match scores for 25 model entries. Its default test split has 1,174 examples.

Released
March 2024
Kind
Electronics
Level
University
Scored
Checked automatically
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Code → Code
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Llama 3.1-405B #163 on the capability index
Last model added
September 2025
Scores refreshed
November 20th, 2026
25
Entries on TuRTLe’s leaderboard, 8 models we track

Qwen3 236B A22B tops the leaderboard with 41.9% Qwen3-Coder-480B-A35B leads with 38.9% among models released by September 2025

Best score per entry from TuRTLe’s leaderboard, read on November 20th, 2026. 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 Qwen3-Coder-480B-A35B, at 38.9%. This view shows only models we track, each at its best.

The newest model with a score here came out in September 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Qwen3 236B A22B 41.9%
  2. 2 Qwen3-Coder-480B-A35B 38.9%
  3. 3 QwenCoder 2.5 14B 37.5%
  4. 4 Qwen2.5-72B 37.4%
  5. 5 Llama 3.1-405B 34.6%
  6. 6 Kimi K2 Sep 2025 33.6%
  7. 7 QwenCoder 2.5 32B 31.1%
  8. 8 Qwen2.5-32B 28.9%
  9. 9 Llama 3.(1-3) 70B 28.7%
  10. 10 Hermes-4-14B 28.4%
  11. 11 QwenCoder 2.5 7B 28.3%
  12. 12 Seed-Coder-8B-Instruct 28.2%
  13. 13 InCoder-32B 27.1%
  14. 14 HaVen-CodeQwen 25.1%
  15. 15 DeepSeek Coder 6.7B 24.6%
  16. 16 DeepSeek Coder 33B 24.6%
  17. 17 CodeLlama 70B 24.3%
  18. 18 CodeV-DS-6.7B 21.3%
  19. 19 CodeV-QW-7B 20.9%
  20. 20 RTLCoder DeepSeek 19.4%
  21. 21 OriGen 17.1%
  22. 22 OpenCoder 8B 16.2%
  23. 23 RTLCoder Mistral 14.8%
  24. 24 StarChat2 15B v0.1 13.4%
  25. 25 CodeV-CL-7B 12.3%

Source: TuRTLe’s leaderboard

#EntryScoreSettingDateSource
1 Qwen3 236B A22B 41.9% Official leaderboard
2 Qwen3-Coder-480B-A35B 38.9% Jul 2025 Official leaderboard
3 QwenCoder 2.5 14B 37.5% Official leaderboard
4 Qwen2.5-72B 37.4% Sep 2024 Official leaderboard
5 Llama 3.1-405B 34.6% Jul 2024 Official leaderboard
6 Kimi K2 Sep 2025 33.6% Sep 2025 Official leaderboard
7 QwenCoder 2.5 32B 31.1% Official leaderboard
8 Qwen2.5-32B 28.9% Sep 2024 Official leaderboard
9 Llama 3.(1-3) 70B 28.7% Official leaderboard
10 Hermes-4-14B 28.4% Official leaderboard
11 QwenCoder 2.5 7B 28.3% Official leaderboard
12 Seed-Coder-8B-Instruct 28.2% May 2025 Official leaderboard
13 InCoder-32B 27.1% Official leaderboard
14 HaVen-CodeQwen 25.1% Official leaderboard
15 DeepSeek Coder 6.7B 24.6% Nov 2023 Official leaderboard
16 DeepSeek Coder 33B 24.6% Nov 2023 Official leaderboard
17 CodeLlama 70B 24.3% Official leaderboard
18 CodeV-DS-6.7B 21.3% Official leaderboard
19 CodeV-QW-7B 20.9% Official leaderboard
20 RTLCoder DeepSeek 19.4% Official leaderboard
21 OriGen 17.1% Official leaderboard
22 OpenCoder 8B 16.2% Official leaderboard
23 RTLCoder Mistral 14.8% Official leaderboard
24 StarChat2 15B v0.1 13.4% Official leaderboard
25 CodeV-CL-7B 12.3% Official leaderboard
#ModelScoreSettingDateSource
1 Qwen3-Coder-480B-A35B 38.9% Jul 2025 Official leaderboard
2 Qwen2.5-72B 37.4% Sep 2024 Official leaderboard
3 Llama 3.1-405B 34.6% Jul 2024 Official leaderboard
4 Kimi K2 Sep 2025 33.6% Sep 2025 Official leaderboard
5 Qwen2.5-32B 28.9% Sep 2024 Official leaderboard
6 Seed-Coder-8B-Instruct 28.2% May 2025 Official leaderboard
7 DeepSeek Coder 6.7B 24.6% Nov 2023 Official leaderboard
8 DeepSeek Coder 33B 24.6% Nov 2023 Official leaderboard

4 models have held number one on TuRTLe’s leaderboard since its release in March 2024

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

DeepSeek Meta Qwen
TuRTLe leaderboard Not tested since Sep 2025

Source: TuRTLe’s 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 TuRTLe’s leaderboard Best so far
0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Qwen3-Coder-480B-A35B Qwen2.5-72B DeepSeek Coder 6.7B 0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Qwen3-Coder-480B-A35B DeepSeek Coder 6.7B

Source: TuRTLe’s 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

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 14 months ago, in September 2025, for Kimi K2. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (25 entries).
  • Pass Models scored 8 models we track, plus 17 other entries on its leaderboard.
  • Warning Newest score The newest came 14 months ago, in September 2025, for Kimi K2.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Llama 3.1-405B (#163).
  • Pass Headroom The best is 38.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on TuRTLe: 8 rows for models we track and 17 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