LiveCodeBench (Vals Public) - Medium
Code · 2024-03-12
The medium difficulty slice of Vals AI's LiveCodeBench public-dataset evaluation surface.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5 | 92.7 |
| Claude Fable 5.1 | 91.9 |
| Qwen3.6 Plus (2026-04-02) | 91.6 |
| Gemini 3.8 Flash | 91.4 |
| DeepSeek-V4-Pro | 91.1 |
| Claude Fable 5 | 90.9 |
| Claude Opus 5 | 90.6 |
| Gemini 3.5 Flash | 90.6 |
| Gemini 3.6 Flash | 90.6 |
| Gemini 3.7 Flash | 90.6 |
| GPT-5.2-Codex | 90.3 |
| Nemotron 3 Ultra 550B A55B (NVFP4) | 90.3 |
| Qwen3.5 Plus (2026-02-15) | 90.1 |
| Opus 4.8 | 89.8 |
| DeepSeek V4 Pro 0813 | 89.8 |
| Gemini 3.1 Pro Preview | 89.8 |
| Grok 4 | 89.8 |
| Inkling | 89.8 |
| Kimi K2.6 | 89.8 |
| Muse Spark 1.1 | 89.8 |
| GPT-5.1 | 89.6 |
| Grok 4.5 | 89.6 |
| Opus 4.6 | 89.3 |
| Opus 4.7 | 89.3 |
| Gemini 3 Pro Preview | 89.3 |
| GPT-5 Nano | 89.3 |
| Inkling-Small | 89.3 |
| o3 | 89.3 |
| O4 Mini | 89.3 |
| Qwen3.7-Max (source unspecified) | 89.3 |
| DeepSeek-V4-Flash-0731 | 89.0 |
| GLM-5.3 Flash | 89.0 |
| GPT-5.1-Codex | 89.0 |
| GPT-5 Mini | 89.0 |
| Grok 4.3 | 89.0 |
| Grok 4.6 | 89.0 |
| Qwen3.8-Max | 88.8 |
| GPT-5.4 Nano | 88.5 |
| Kimi K3 | 88.5 |
| Qwen3.5-Flash (source unspecified) | 88.5 |