Noometry

Coding benchmark

FrontierSWE leaderboard

As of October 2026, GPT-6 Astra has the highest published FrontierSWE score on Noometry at 65.5%, out of 18 models with results.

Last verified

About FrontierSWE

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

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

Top 15 models

  1. GPT-6 Astra 65.5%
  2. Claude Opus 5.5 62.3%
  3. Claude Sonnet 5.5 61.9%
  4. Claude Fable 5.1 56.3%
  5. Gemini 4 Argon 55%
  6. Claude Opus 5 52%
  7. Claude Fable 5 47%
  8. GPT-5.6 Sol 32.2%
  9. GLM-5.3 30.2%
  10. Grok 4.7 29.5%
  11. Kimi K3 25.9%
  12. Grok 4.6 25.3%
  13. Gemini 3.7 Flash 20.3%
  14. Gemini 3.8 Flash 19.6%
  15. GLM-5.3-Flash 18.1%

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

All results

#ModelProviderScoreSettingSourceDate
1GPT-6 Astra OpenAI65.5%maxEpoch AI
2Claude Opus 5.5 Anthropic62.3%maxEpoch AI
3Claude Sonnet 5.5 Anthropic61.9%maxEpoch AI
4Claude Fable 5.1 Anthropic56.3%maxEpoch AI
5Gemini 4 Argon Google55%highEpoch AI
6Claude Opus 5 Anthropic52%maxEpoch AI
7Claude Fable 5 Anthropic47%maxEpoch AI
8GPT-5.6 Sol OpenAI32.2%maxEpoch AI
9GLM-5.3 Z.ai (Zhipu)30.2%maxEpoch AI
10Grok 4.7 xAI29.5%xhighEpoch AI
11Kimi K3 Moonshot AI25.9%maxEpoch AI
12Grok 4.6 xAI25.3%xhighEpoch AI
13Gemini 3.7 Flash Google20.3%highEpoch AI
14Gemini 3.8 Flash Google19.6%highEpoch AI
15GLM-5.3-Flash Z.ai (Zhipu)18.1%maxEpoch AI
16Qwen3.8 Max Alibaba (Qwen)17.8%xhighEpoch AI
17Muse Spark 1.2 Meta12%xhighEpoch AI
18Inkling Thinking Machines Lab4.1%xhighEpoch AI

Compare the leaders

Other coding benchmarks

Frequently asked questions

Which model has the highest FrontierSWE score?

As of October 2026, GPT-6 Astra has the highest published FrontierSWE score on Noometry at 65.5%, out of 18 models with results.

What is the best open-weight model on FrontierSWE?

GLM-5.3 has the highest FrontierSWE accuracy among open-weight models at 30.2%, ranking 9 of 18 overall.