LiveBench — leaderboard
Continuously updated benchmark measuring many capabilities. Sort models by weighted score or sub-task scores. Tests math, coding, reasoning, language, instruction following, and data analysis.
Metric: LiveBench average (self-reported). Source: benchmarklist.com. Status: saturation imminent. 43 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 (xHigh) | 81.28 |
| 2 | GPT-5.4 (xHigh) | 80.91 |
| 3 | Gemini 3.1 Pro (Preview) (High) | 80.71 |
| 4 | Claude Opus 4.8 (xHigh) | 77.55 |
| 5 | Claude Opus 4.7 (xHigh) | 77.1 |
| 6 | Claude Opus 4.6 (Thinking, High) | 76.79 |
| 7 | Claude Opus 4.8 (High) | 76.18 |
| 8 | Gemini 3.5 Flash (High) | 75.79 |
| 9 | Claude Fable 5 (High) | 75.74 |
| 10 | Claude Sonnet 4.6 (Thinking, High) | 75.59 |
| 11 | GPT-5.2 (High) | 75.38 |
| 12 | Qwen 3.7 Max (Max) | 75.15 |
| 13 | Claude Opus 4.7 (High) | 74.66 |
| 14 | DeepSeek V4 Pro | 74.39 |
| 15 | GPT-5.2 Codex | 74.33 |
Interactive version: theaggregate.ai/benchmark?slug=livebench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.