Benchmark Almanac

VeriGen design problems

Gemini 2.5 Flash Sep 2025 77.7% leads since September 2025
Not in model cards

Models complete Verilog modules for 17 digital circuits. Simulation checks whether the generated code satisfies each design’s testbench.

TuRTLe publishes functional scores for 45 model entries using the medium-detail prompts.

Released
December 2022
Kind
Electronics
Level
University
Scored
Checked automatically
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Code → Code
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GLM-5.2 #55 on the capability index
Last model added
June 2026
Scores refreshed
November 20th, 2026
45
Entries on TuRTLe’s leaderboard, 27 models we track

Gemini 2.5 Flash leads with 77.7% among models released by June 2026 Gemini 2.5 Flash leads with 77.7% among models released by June 2026

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. This view shows only models we track, each at its best.

The newest model with a score here came out in June 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 Sep 2025, medium 77.7%
  2. 2 OriGen 76.5%
  3. 3 Qwen3-Coder-480B-A35B 75.3%
  4. 4 DeepSeek-V3.1-Terminus 75.3%
  5. 5 DeepSeek-R1 74.1%
  6. 6 DeepSeek-R1 May 2025 74.1%
  7. 7 Kimi K2.5 74.1%
  8. 8 seed-oss-36b-instruct 70.6%
  9. 9 Kimi K2 Sep 2025 69.4%
  10. 10 gpt-oss-120b 68.2%
  11. 11 CodeV R1 Distill Qwen 7B 65.9%
  12. 12 Llama 3.3 70B Instruct 65.9%
  13. 13 StarChat2 15B v0.1 63.5%
  14. 14 GLM-5 FP8 63.5%
  15. 15 HaVen-CodeQwen 62.4%
  16. 16 SeedCoder 8B Reasoning 62.4%
  17. 17 Qwen3-235B-A22B 62.4%
  18. 18 CodeV-DS-6.7B 61.2%
  19. 19 CodeV-QW-7B 60%
  20. 20 InCoder-32B-Thinking 60%
  21. 21 Llama 3.1-405B 58.8%
  22. 22 Hermes-4-14B-Reasoning 56.5%
  23. 23 RTLCoder DeepSeek 55.3%
  24. 24 GLM-5.2 FP8 55.3%
  25. 25 Qwen2.5-Coder-32B-Instruct 54.1%
  26. 26 QwQ-32B 52.9%
  27. 27 Gemma 4 31B IT 52.9%
  28. 28 RTLCoder Mistral 51.8%
  29. 29 Magistral Small 1.0 50.6%
  30. 30 gpt-oss-20b 50.6%
  31. 31 CodeV-CL-7B 49.4%
  32. 32 InCoder-32B 49.4%
  33. 33 Qwen2.5-Coder-14B-Instruct 48.2%
  34. 34 CodeLlama 70B 47.1%
  35. 35 Qwen2.5-72B 47.1%
  36. 36 Hermes-4-14B 45.9%
  37. 37 OpenCoder 8B 45.9%
  38. 38 DeepSeek Coder 33B 45.9%
  39. 39 Qwen2.5-32B 45.9%
  40. 40 Seed-Coder-8B-Instruct 42.4%
  41. 41 QwenCoder 2.5 7B 40%
  42. 42 Qwen3-8B 40%
  43. 43 DeepSeek Coder 6.7B 37.6%
  44. 44 DeepCoder 14B 32.9%
  45. 45 DeepSeek-R1-Distill-Qwen-14B 28.2%

Source: TuRTLe’s leaderboard

#EntryScoreSettingDateSource
1 Gemini 2.5 Flash Sep 2025 77.7% medium Sep 2025 Official leaderboard
2 OriGen 76.5% Official leaderboard
3 Qwen3-Coder-480B-A35B 75.3% Jul 2025 Official leaderboard
4 DeepSeek-V3.1-Terminus 75.3% Sep 2025 Official leaderboard
5 DeepSeek-R1 74.1% Jan 2025 Official leaderboard
6 DeepSeek-R1 May 2025 74.1% May 2025 Official leaderboard
7 Kimi K2.5 74.1% Jan 2026 Official leaderboard
8 seed-oss-36b-instruct 70.6% Aug 2025 Official leaderboard
9 Kimi K2 Sep 2025 69.4% Sep 2025 Official leaderboard
10 gpt-oss-120b 68.2% Aug 2025 Official leaderboard
11 CodeV R1 Distill Qwen 7B 65.9% Official leaderboard
12 Llama 3.3 70B Instruct 65.9% Dec 2024 Official leaderboard
13 StarChat2 15B v0.1 63.5% Official leaderboard
14 GLM-5 63.5% FP8 Feb 2026 Official leaderboard
15 HaVen-CodeQwen 62.4% Official leaderboard
16 SeedCoder 8B Reasoning 62.4% Official leaderboard
17 Qwen3-235B-A22B 62.4% Apr 2025 Official leaderboard
18 CodeV-DS-6.7B 61.2% Official leaderboard
19 CodeV-QW-7B 60% Official leaderboard
20 InCoder-32B-Thinking 60% Official leaderboard
21 Llama 3.1-405B 58.8% Jul 2024 Official leaderboard
22 Hermes-4-14B-Reasoning 56.5% Official leaderboard
23 RTLCoder DeepSeek 55.3% Official leaderboard
24 GLM-5.2 55.3% FP8 Jun 2026 Official leaderboard
25 Qwen2.5-Coder-32B-Instruct 54.1% Nov 2024 Official leaderboard
26 QwQ-32B 52.9% Mar 2025 Official leaderboard
27 Gemma 4 31B IT 52.9% Apr 2026 Official leaderboard
28 RTLCoder Mistral 51.8% Official leaderboard
29 Magistral Small 1.0 50.6% Jun 2025 Official leaderboard
30 gpt-oss-20b 50.6% Aug 2025 Official leaderboard
31 CodeV-CL-7B 49.4% Official leaderboard
32 InCoder-32B 49.4% Official leaderboard
33 Qwen2.5-Coder-14B-Instruct 48.2% Nov 2024 Official leaderboard
34 CodeLlama 70B 47.1% Official leaderboard
35 Qwen2.5-72B 47.1% Sep 2024 Official leaderboard
36 Hermes-4-14B 45.9% Official leaderboard
37 OpenCoder 8B 45.9% Official leaderboard
38 DeepSeek Coder 33B 45.9% Nov 2023 Official leaderboard
39 Qwen2.5-32B 45.9% Sep 2024 Official leaderboard
40 Seed-Coder-8B-Instruct 42.4% May 2025 Official leaderboard
41 QwenCoder 2.5 7B 40% Official leaderboard
42 Qwen3-8B 40% Apr 2025 Official leaderboard
43 DeepSeek Coder 6.7B 37.6% Nov 2023 Official leaderboard
44 DeepCoder 14B 32.9% Official leaderboard
45 DeepSeek-R1-Distill-Qwen-14B 28.2% Jan 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 2.5 Flash Sep 2025 77.7% medium Sep 2025 Official leaderboard
2 Qwen3-Coder-480B-A35B 75.3% Jul 2025 Official leaderboard
3 DeepSeek-V3.1-Terminus 75.3% Sep 2025 Official leaderboard
4 DeepSeek-R1 74.1% Jan 2025 Official leaderboard
5 DeepSeek-R1 May 2025 74.1% May 2025 Official leaderboard
6 Kimi K2.5 74.1% Jan 2026 Official leaderboard
7 seed-oss-36b-instruct 70.6% Aug 2025 Official leaderboard
8 Kimi K2 Sep 2025 69.4% Sep 2025 Official leaderboard
9 gpt-oss-120b 68.2% Aug 2025 Official leaderboard
10 Llama 3.3 70B Instruct 65.9% Dec 2024 Official leaderboard
11 GLM-5 63.5% FP8 Feb 2026 Official leaderboard
12 Qwen3-235B-A22B 62.4% Apr 2025 Official leaderboard
13 Llama 3.1-405B 58.8% Jul 2024 Official leaderboard
14 GLM-5.2 55.3% FP8 Jun 2026 Official leaderboard
15 Qwen2.5-Coder-32B-Instruct 54.1% Nov 2024 Official leaderboard
16 QwQ-32B 52.9% Mar 2025 Official leaderboard
17 Gemma 4 31B IT 52.9% Apr 2026 Official leaderboard
18 Magistral Small 1.0 50.6% Jun 2025 Official leaderboard
19 gpt-oss-20b 50.6% Aug 2025 Official leaderboard
20 Qwen2.5-Coder-14B-Instruct 48.2% Nov 2024 Official leaderboard
21 Qwen2.5-72B 47.1% Sep 2024 Official leaderboard
22 DeepSeek Coder 33B 45.9% Nov 2023 Official leaderboard
23 Qwen2.5-32B 45.9% Sep 2024 Official leaderboard
24 Seed-Coder-8B-Instruct 42.4% May 2025 Official leaderboard
25 Qwen3-8B 40% Apr 2025 Official leaderboard
26 DeepSeek Coder 6.7B 37.6% Nov 2023 Official leaderboard
27 DeepSeek-R1-Distill-Qwen-14B 28.2% Jan 2025 Official leaderboard

6 models have held number one on TuRTLe’s leaderboard since 2023

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

DeepSeek Meta Qwen Google DeepMind
TuRTLe leaderboard Not tested since Jun 2026

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 2026 Near the ceiling (85%) Gemini 2.5 Flash DeepSeek-R1 Llama 3.3 70B Instruct Llama 3.1-405B DeepSeek Coder 33B 0% 20% 40% 60% 80% 100% 2024 2025 2026 Near the ceiling (85%) Gemini 2.5 Flash Llama 3.3 70B Instruct

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

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (45 entries).
  • Pass Models scored 27 models we track, plus 18 other entries on its leaderboard.
  • Pass Newest score The newest is from June 2026, for GLM-5.2.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GLM-5.2 (#55).
  • Pass Headroom The best is 77.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on TuRTLe: 27 rows for models we track and 18 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