ALL Bench LLM: leaderboard
Composite LLM leaderboard aggregating cross-verified scores across reasoning, knowledge, coding, and instruction-following evaluations.
Metric: Average Numeric Benchmark Score (%). Source: huggingface.co. Saturation forecast: Not forecast. 39 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.1 Fast | 81.77 | #217 |
| 2 | Step 3.5 Flash | 77.28 | #270 |
| 3 | Claude Sonnet 4.6 | 75.9 | #90 |
| 4 | Qwen 3.5 Flash | 75 | #232 |
| 5 | Solar Open 100B | 75 | #406 |
| 6 | Qwen3-Next-80B | 74.75 | #312 |
| 7 | GPT-5.3 Codex | 74.67 | #69 |
| 8 | Gemini 3.1 Pro (Preview) | 74.62 | #58 |
| 9 | Qwen 3.5 397B A17B | 73.94 | #146 |
| 10 | Claude Opus 4.6 | 73.69 | #63 |
| 11 | GPT-5.2 | 73.02 | #110 |
| 12 | K-EXAONE | 72.78 | #378 |
| 13 | Gemini 3 Flash | 72.09 | #99 |
| 14 | Claude Sonnet 4.5 | 72.07 | #142 |
| 15 | Qwen 3.5 4B | 71.22 | #475 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=all-bench-llm · How It Works · Data refreshed daily, snapshot 2026-10-20.