Capability
Indexes · 2026-10-28
Synthetic general capability score from the Atlas factor model. Higher scores indicate stronger estimated overall benchmark performance.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5.5 | 168.2 |
| Gemini 4 Carbon | 168.1 |
| GPT-6 Astra | 168.0 |
| Claude Sonnet 5.5 | 167.1 |
| Gemini 4 Argon | 167.0 |
| GPT-6.1 Sol | 166.9 |
| Claude Fable 5.1 | 166.5 |
| Kimi K3.1 Pro | 165.7 |
| GPT-6.1 Terra | 165.5 |
| Muse Wonder Full | 165.4 |
| Claude Fable 5 | 165.3 |
| Claude Opus 5 | 164.9 |
| GPT-6 Sol | 164.4 |
| Muse Wonder | 164.4 |
| Claude Mythos Preview | 164.2 |
| GPT-5.6 Sol Pro | 164.0 |
| GPT-5.6 Sol | 163.6 |
| GPT-5.5 Pro | 163.1 |
| Gemini 3 Deep Think | 162.9 |
| Fugu Ultra v1.0 | 162.8 |
| GPT-5.6 Terra Pro | 162.2 |
| Claude Haiku 5.5 | 162.2 |
| DeepSeek V4.1 Flash 1017 | 162.2 |
| Grok 4.7 | 162.0 |
| Muse Spark 1.3 | 161.8 |
| GPT-5.4 Pro | 161.6 |
| Grok 4.6 | 161.4 |
| Gemini 3.8 Flash | 161.3 |
| GPT-5.6 Terra | 161.3 |
| MiMo-V2.6-Pro | 161.3 |
| Qwen3.8 Max (0902) | 161.3 |
| GLM-5.3 | 161.1 |
| MiniMax M3.1 Flash | 161.1 |
| Atria Dawn Preview | 161.0 |
| GPT-5.5 | 160.7 |
| Ling 3.1 Flash 1019 | 160.7 |
| Opus 4.8 | 160.6 |
| GPT-5.6 Luna Pro | 160.5 |
| Ling 3.1 Flash | 160.5 |
| Ember-1 | 160.5 |