ReLE - Reasoning - BBH: leaderboard
ReLE Chinese LLM benchmark (NoneLinear, formerly CLiB), a continuously updated Chinese-language evaluation run through each model's API. Reasoning sub-dimension: ReLE's own run of BIG-Bench Hard symbolic reasoning tasks; not a copy of other BBH leaderboards. Accuracy 0-100, higher is better.
Metric: Accuracy (%). Source: nonelinear.com. Saturation forecast: Around March 2027. 148 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 32B | 89.9 | #449 |
| 2 | Gemini 3.8 Flash | 87.8 | #30 |
| 3 | Gemini 3.1 Pro (Preview) | 87.5 | #64 |
| 4 | Gemini 4 Argon | 87.5 | #9 |
| 5 | Claude Opus 4.6 | 86.8 | #67 |
| 6 | Gemini 3.5 Flash | 86 | #66 |
| 7 | Claude Opus 5 | 86 | #16 |
| 8 | Kimi K3.1 Pro | 86 | #17 |
| 9 | Gemini 3.7 Flash | 85.9 | #24 |
| 10 | GPT-4o | 85.7 | #357 |
| 11 | DeepSeek R1 0528 | 85.7 | #232 |
| 12 | Qwen 3.7 Plus | 85.7 | #98 |
| 13 | Claude Opus 4.8 (Thinking) | 85.7 | #39 (Claude Opus 4.8) |
| 14 | Claude Sonnet 5 (Thinking) | 85.7 | #65 (Claude Sonnet 5) |
| 15 | GLM-5.3 | 85.6 | #44 |
No result here: #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=rele-reasoning-bbh · How It Works · Data refreshed daily, snapshot 2026-11-05.