BLXBench: leaderboard
Community benchmark runner and public leaderboard for AI model performance across coding, debugging, reasoning, hallucination, refactoring, security, and speed slices.
Metric: Score (self-reported). Source: benchmarklist.com. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok 4.3 | 85.5 |
| 2 | Claude Opus 4.7 | 84.8 |
| 3 | Qwen 3.6 Flash | 82.8 |
| 4 | Grok 4.20 | 79.1 |
| 5 | Ling-2.6-1T | 75.3 |
| 6 | Mistral Small 4 | 75.2 |
| 7 | Claude Opus 4.6 | 71.1 |
| 8 | GPT-5.5 | 65.9 |
| 9 | Granite 4.1 8B | 64.6 |
| 10 | DeepSeek V4 Flash | 48.3 |
| 11 | GPT-5.5 Pro | 44.5 |
| 12 | MiMo-V2.5-Pro | 44 |
| 13 | Qwen 3.6 35B A3B | 33.9 |
| 14 | Qwen 3.6 27B | 29 |
| 15 | MiMo-V2.5 | 15.7 |
Interactive version: theaggregate.ai/benchmark?slug=blxbench · How It Works · Data refreshed daily, snapshot 2026-09-05.