Vals AI LiveCodeBench — leaderboard
Metric: Accuracy (%). Source: www.vals.ai. 128 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 | 89.78 |
| 2 | Gemini 3.1 Pro (Preview) | 88.48 |
| 3 | GPT-5.2 Codex | 87.99 |
| 4 | Claude Opus 4.8 | 87.82 |
| 5 | Gemini 3.5 Flash | 87.6 |
| 6 | DeepSeek V4 Pro | 87.48 |
| 7 | Grok 4.5 | 87.35 |
| 8 | GPT-5.3 Codex | 87.31 |
| 9 | Kimi K3 | 87.19 |
| 10 | Qwen 3.7 Max | 87.06 |
| 11 | Kimi K2.6 | 86.77 |
| 12 | GPT-5 Mini | 86.61 |
| 13 | GPT-5.1 | 86.49 |
| 14 | Gemini 3 Pro (Preview) | 86.41 |
| 15 | Nemotron 3 Ultra | 85.98 |
Interactive version: theaggregate.ai/benchmark?slug=vals-ai-livecodebench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.