# LLM Stats

A website that gathers public test results, prices and speeds for hundreds of AI models and ranks them with a single combined score.

Category: aggregators. Type: leaderboard. Status: active.
Metric: LLM Stats composite score (higher is better) from public benchmark results (higher is better).
Source: LLM Stats (llm-stats.com) — https://llm-stats.com/, retrieved 2026-11-17, license: not stated.

| Rank | Model | Organization | LLM Stats score | Reasoning | Coding | Agent |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.2 | Anthropic | 60.8 | 60.0 | 49.8 | 46.3 |
| 2 | Claude Mythos 5.2 | Anthropic | 59.3 | 53.4 | 44.5 | 44.2 |
| 3 | Claude Opus 5.5 | Anthropic | 58.5 | 57.4 | 47.8 | 43.6 |
| 4 | Gemini 4 Carbon | Google | 57.4 | 54.0 | 45.7 | 43.7 |
| 5 | Claude Sonnet 5.5 | Anthropic | 56.7 | 55.4 | 46.0 | 41.9 |
| 6 | GPT-6 Astra | OpenAI | 56.3 | 55.0 | 43.9 | 42.3 |
| 7 | Kimi K3.1 Pro | Moonshot AI | 54.4 | 53.3 | 40.5 | 39.9 |
| 8 | Claude Mythos Preview | Anthropic | 53.6 | 54.6 | 42.5 | 35.1 |
| 9 | Gemini 4 Argon | Google | 53.3 | 51.2 | 41.3 | 39.6 |
| 10 | Muse Wonder Full | Meta | 52.6 | 44.8 | 42.0 | 40.4 |
| 11 | Claude Opus 5 | Anthropic | 52.5 | 52.5 | 37.3 | 36.7 |
| 12 | GPT-6.1 Sol | OpenAI | 52.2 | 49.6 | 37.6 | 35.8 |
| 13 | Claude Fable 5.1 | Anthropic | 52.0 | 49.8 | 34.4 | 37.1 |
| 14 | Muse Wonder Full | Meta | 52.0 | 40.1 | 40.9 | 40.0 |
| 15 | Claude Fable 5 | Anthropic | 51.9 | 49.8 | 41.6 | 37.0 |

Notes: Homepage leaderboard table (page says 'updated continuously'); includes some unreleased/preview models. Scores are an aggregate of other benchmarks' public results.
