OpenCompass Research - LiveCodeBench v6: leaderboard
OpenCompass research-table score on LiveCodeBench v6, evaluating coding ability on recent programming problems.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Around January 2027. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5 | 86.2 | #141 |
| 2 | Step 3.5 Flash | 83.9 | #270 |
| 3 | GLM-4.7 | 83.8 | #192 |
| 4 | GPT-5 (2025-08-07) (High) | 83.2 | |
| 5 | Qwen 3.5 397B A17B | 83 | #146 |
| 6 | Gemini 3 Pro (Preview) | 82.9 | #66 |
| 7 | DeepSeek V3.2 Speciale | 80.9 | #128 |
| 8 | GPT-5 Mini (2025-08-07) (High) | 80.8 | #172 (GPT-5 Mini (2025-08-07)) |
| 9 | Kimi K2.5 | 80.6 | #144 |
| 10 | GPT-OSS-120B (High) | 78.4 | #341 (GPT-OSS-120B) |
| 11 | GLM-4.6 | 78.2 | #252 |
| 12 | Kimi K2 (Thinking) | 77.1 | #243 (Kimi K2) |
| 13 | DeepSeek V3.2 | 75.4 | #205 |
| 14 | MiniMax-M2 | 74 | #317 |
| 15 | GPT-5 Nano (2025-08-07) (High) | 74 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=opencompass-research-livecodebench-v6 · How It Works · Data refreshed daily, snapshot 2026-10-20.