All tests
Overall
LiveBench overall
Broad language-model performance across the current LiveBench release.
Published by LiveBench under Apache-2.0. Compare it with other tests
Top score83.4%Claude Fable 5.1 (max effort), Anthropic
Models tested63from 11 makers
Median score75.8%Half the models score above it
Scale0 to 100%The share of tasks done right
Top 15
Each model at its best configuration, coloured by maker.
The best model from each maker
Each maker’s highest-scoring model on this test, and its rank.
| Maker | Best model | Score | Rank |
|---|---|---|---|
 Anthropic | Claude Fable 5.1 (max effort) | 83.4% | #1 of 63 |
 OpenAI | GPT-6 Astra (max) | 82.2% | #4 of 63 |
 DeepSeek | DeepSeek V4.1 Flash (max) | 81.1% | #7 of 63 |
 Moonshot AI | Kimi K3 | 79.2% | #13 of 63 |
 Google | Gemini 3.7 Flash (high) | 78.8% | #14 of 63 |
 Alibaba | Qwen3.8 Max | 78.5% | #15 of 63 |
 xAI | Grok 4.6 | 78.0% | #16 of 63 |
 Z.ai | GLM-5.3 | 76.1% | #30 of 63 |
 NVIDIA | Nemotron 3 Ultra 550B A55B | 67.4% | #57 of 63 |
 MiniMax | MiniMax-M3 | 67.3% | #58 of 63 |
Every result
63 models, each once at its best configuration, highest score first. Models with the same score share a rank.
| Rank | Model | Maker | Score |
|---|---|---|---|
| #1 | Claude Fable 5.1 (max effort) | Anthropic | 83.4% |
| #2 | Claude Opus 5.5 (max effort) | Anthropic | 83.2% |
| #3 | Claude Fable 5 (max effort) | Anthropic | 83.0% |
| #4 | GPT-6 Astra (max) | OpenAI | 82.2% |
| #5 | GPT-6.1 Sol (max) | OpenAI | 81.6% |
| #6 | Muse Spark 1.3 (xhigh) |  Meta AI | 81.6% |
| #7 | DeepSeek V4.1 Flash (max) | DeepSeek | 81.1% |
| #8 | GPT-5.6 Sol (max) | OpenAI | 81.0% |
| #9 | GPT-5.5 (xhigh) | OpenAI | 80.2% |
| #10 | Claude Opus 5 (max effort) | Anthropic | 80.1% |
| #11 | Smaug Agentic | Not reported | 79.5% |
| #12 | GPT-6 Sol (max) | OpenAI | 79.3% |
| #13 | Kimi K3 | Moonshot AI | 79.2% |
| #14 | Gemini 3.7 Flash (high) | Google | 78.8% |
| #15 | Qwen3.8 Max | Alibaba | 78.5% |
| #16 | Grok 4.6 | xAI | 78.0% |
| #17 | GPT-5.4 (xhigh) | OpenAI | 78.0% |
| #18 | Muse Spark 1.2 (xhigh) | Meta AI | 78.0% |
| #19 | GPT-5.6 Terra (max) | OpenAI | 77.9% |
| #20 | Claude Sonnet 5.5 (xhigh effort) | Anthropic | 77.8% |
| #21 | DeepSeek V4 Pro 0813 | DeepSeek | 77.4% |
| #21 | Smaug Flash | Not reported | 77.4% |
| #23 | Grok 4.7 (xhigh) | xAI | 77.4% |
| #24 | Gemini 3.1 Pro | Google | 77.0% |
| #25 | Smaug Mini | Not reported | 76.9% |
How it is scored
Equal-weight mean of all 7 category means, matching the published aggregation method for complete results. This is LiveBench performance, not a universal best-model score.
Results from LiveBench (Apache-2.0), as published. Model names link to their pages where the site can match them; makers come from the model’s page, or else from the publisher.