Noometry

Reasoning benchmark

Surface Evolver Bench leaderboard

As of October 2026, Claude Fable 5 has the highest published Surface Evolver Bench score on Noometry at 95%, out of 25 models with results.

Last verified

About Surface Evolver Bench

A description with primary sources is being prepared for this benchmark.

Category
Reasoning
Introduced
2026
Unit
Percent (random guessing ≈ 0%)
Official site
epoch.ai

Top 15 models

  1. Claude Fable 5 95%
  2. Kimi K3 95%
  3. GPT-5.6 Sol 93.1%
  4. GPT-5.5 88.1%
  5. Claude Opus 4.8 87.5%
  6. GPT-5.6 Terra 83.8%
  7. Gemini 3.8 Flash 76.9%
  8. Grok 4.5 74.4%
  9. GPT-5.6 Luna 61.9%
  10. Claude Sonnet 5 60%
  11. Gemini 3.5 Flash 58.1%
  12. GLM-5.2 55.6%
  13. MiniMax-M3 55%
  14. GLM-5.3-Flash 52.5%
  15. Muse Spark 1.1 52.5%

Sponsored placements are available on pages like this one. Advertise on Noometry

All results

#ModelProviderScoreSettingSourceDate
1Claude Fable 5 Anthropic95%highEpoch AI
2Kimi K3 Moonshot AI95%Epoch AI
3GPT-5.6 Sol OpenAI93.1%xhighEpoch AI
4GPT-5.5 OpenAI88.1%highEpoch AI
5Claude Opus 4.8 Anthropic87.5%highEpoch AI
6GPT-5.6 Terra OpenAI83.8%xhighEpoch AI
7Gemini 3.8 Flash Google76.9%highEpoch AI
8Grok 4.5 xAI74.4%highEpoch AI
9GPT-5.6 Luna OpenAI61.9%mediumEpoch AI
10Claude Sonnet 5 Anthropic60%mediumEpoch AI
11Gemini 3.5 Flash Google58.1%mediumEpoch AI
12GLM-5.2 Z.ai (Zhipu)55.6%highEpoch AI
13MiniMax-M3 MiniMax55%Epoch AI
14GLM-5.3-Flash Z.ai (Zhipu)52.5%maxEpoch AI
15Muse Spark 1.1 Meta52.5%highEpoch AI
16Kimi K2.7 Code Moonshot AI48.8%Epoch AI
17DeepSeek V4.1 Flash DeepSeek46.3%highEpoch AI
18Qwen3.8 27B Alibaba (Qwen)45%xhighEpoch AI
19Qwen3.6 35B-A3B Alibaba (Qwen)44.4%Epoch AI
20DeepSeek V4 Pro DeepSeek40%highEpoch AI
21Gemma 4 31B IT Google30.6%Epoch AI
22Mistral Medium Mistral AI26.9%Epoch AI
23gpt-oss-120b OpenAI25%Epoch AI
24Laguna M.1 Poolside15.6%Epoch AI
25Trinity Large Thinking Arcee AI15.6%Epoch AI

Compare the leaders

Other reasoning benchmarks

Frequently asked questions

Which model has the highest Surface Evolver Bench score?

As of October 2026, Claude Fable 5 has the highest published Surface Evolver Bench score on Noometry at 95%, out of 25 models with results.

What is the best open-weight model on Surface Evolver Bench?

Kimi K3 has the highest Surface Evolver Bench accuracy among open-weight models at 95%, ranking 2 of 25 overall.