LLM benchmarks
Every model's score on every benchmark we track, each linked to where it was published. Independent runs are listed apart from the numbers labs report for their own models. Many of them feed the Respan Index.
Every model's score on every benchmark we track, each linked to where it was published. Independent runs are listed apart from the numbers labs report for their own models. Many of them feed the Respan Index.
Every model's score on every benchmark we track, each linked to where it was published. Independent runs are listed apart from the numbers labs report for their own models. Many of them feed the Respan Index.
Every model's score on every benchmark we track, each linked to where it was published. Independent runs are listed apart from the numbers labs report for their own models. Many of them feed the Respan Index.
Models ranked on one 0 to 100 score from public benchmarks and lab results.
173 ranked modelsVersion 1.1How it is scored
| Model | Respan Index | |
|---|---|---|
| 1. | Claude Fable 5.2by Anthropic | 80.9 |
| 2. | GPT-6.2 Astraby OpenAI | 76.5 |
| 3. | Gemini 4 Carbonby Google | 76.1 |
| 4. | Claude Opus 5.5by Anthropic | 76.0 |
| 5. | Claude Sonnet 5.5by Anthropic | 73.3 |
| 6. | Gemini 4 Hydrogenby Google | 73.0 |
| 7. | Claude Fable 5.1by Anthropic | 72.4 |
| 8. | GPT-6 Astraby OpenAI | 72.3 |
| 9. | Kimi K3.1 Proby Moonshot AI | 71.5 |
| 10. | Muse Wonder Fullby Meta | 70.9 |
| 11. | GPT-6.1 Solby OpenAI | 69.8 |
| 12. | Gemini 4 Argonby Google | 69.3 |
| 13. | GPT-6.1 Terraby OpenAI | 69.1 |
| 14. | Claude Opus 5by Anthropic | 67.0 |
| 15. | Claude Fable 5by Anthropic | 63.9 |
| 16. | GPT-6 Solby OpenAI | 63.0 |
| 17. | GPT-5.6 Solby OpenAI | 61.7 |
| 18. | Muse Wonderby Meta | 61.0 |
| 19. | Muse Emberby Meta | 59.7 |
| 20. | DeepSeek-V4.1-Flash-1017by DeepSeek | 59.4 |
| 21. | Claude Haiku 5.5by Anthropic | 56.8 |
| 22. | MiMo-V2.6-Proby Xiaomi MiMo | 56.1 |
| 23. | Muse Spark 1.3by Meta | 55.8 |
| 24. | Claude Opus 4.8by Anthropic | 55.7 |
| 24. | Step 5 Previewby StepFun | 55.7 |
* Estimated, no published scores in that category.
A human-preference leaderboard: people compare two anonymous models' answers blind and vote, producing an Elo rating from millions of pairwise votes.
Arena184 modelsRun by huggingface.co/lmarena-ai, Nov 11, 2026