BLXBench: leaderboard
Community benchmark runner and public leaderboard for AI model performance across coding, debugging, reasoning, hallucination, refactoring, security, and speed slices.
Metric: Score (imported). Source: blxbench.com. Source: benchmarklist.com. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.3 | 85.5 | #188 |
| 2 | Claude Opus 4.7 | 84.8 | #49 |
| 3 | Qwen 3.6 Flash | 82.8 | #214 |
| 4 | Grok 4.20 | 79.1 | #175 |
| 5 | Ling-2.6-1T | 75.3 | #309 |
| 6 | Mistral Small 4 | 75.2 | #517 |
| 7 | Claude Opus 4.6 | 71.1 | #63 |
| 8 | GPT-5.5 | 65.9 | #31 |
| 9 | Granite 4.1 8B | 64.6 | #930 |
| 10 | DeepSeek V4 Flash | 48.3 | #137 |
| 11 | GPT-5.5 Pro | 44.5 | #17 |
| 12 | MiMo-V2.5-Pro | 44 | #158 |
| 13 | Qwen 3.6 35B A3B | 33.9 | #226 |
| 14 | Nemotron 3 Nano Omni 30B-A3B (Reasoning) | 29.9 | |
| 15 | Qwen 3.6 27B | 29 | #170 |
Interactive version: theaggregate.ai/benchmark?slug=blxbench · How It Works · Data refreshed daily, snapshot 2026-10-20.