BenchLM: leaderboard
Composite LLM leaderboard aggregating current model performance across agentic, coding, reasoning, grounded multimodal, knowledge, multilingual, instruction-following, and math categories.
Metric: Overall Score. Source: benchlm.ai. Status: years away from saturation. 200 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.1 | 83 |
| 2 | GPT-6 | 81 |
| 3 | Claude Fable 5 | 80.9 |
| 4 | Claude Opus 5 | 80.7 |
| 5 | GPT-5.6 Sol | 79.7 |
| 6 | Gemini 3.8 Flash | 78.4 |
| 7 | Kimi K3 | 74.9 |
| 8 | GPT-5.5 | 73.3 |
| 9 | Qwen 3.8 Max | 72.4 |
| 10 | Claude Opus 4.8 | 72.3 |
| 11 | Muse Spark 1.2 | 71.9 |
| 12 | Muse Spark 1.1 | 71.8 |
| 13 | GPT-5.6 Terra | 71.4 |
| 14 | GPT-5.4 | 71 |
| 15 | Claude Sonnet 5 | 70.8 |
Interactive version: theaggregate.ai/benchmark?slug=benchlm · How It Works · Data refreshed daily, snapshot 2026-09-05.