LiveBench Research: leaderboard
Metric: Score. Source: livebench.ai. Saturation forecast: Around March 2028. 36 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 | 88.9 | #12 |
| 2 | Claude Fable 5 | 86.9 | #15 |
| 3 | Grok 4.7 | 86.4 | #26 |
| 4 | Grok 4.5 | 85.9 | #48 |
| 5 | DeepSeek V4.1 Flash | 85.3 | #39 |
| 6 | GPT-6 Astra | 85.2 | #5 |
| 7 | Claude Opus 5.5 | 85 | #3 |
| 8 | Grok 4.6 | 84.5 | #27 |
| 9 | Claude Sonnet 5.5 | 83.6 | #11 |
| 10 | GPT-6.1 Sol | 83.5 | #9 |
| 11 | Gemini 4 Argon | 83.1 | #7 |
| 12 | Claude Sonnet 5 | 82.8 | #56 |
| 13 | Claude Fable 5.1 | 82 | #8 |
| 14 | Claude Opus 4.8 | 81.8 | #36 |
| 15 | Gemini 3.8 Flash | 79.1 | #21 |
No result here: #10 Muse Wonder Full.
Interactive version: theaggregate.ai/benchmark?slug=livebench-research · How It Works · Data refreshed daily, snapshot 2026-10-14.