Every number
Full score matrix
Filter, search, and sort. Click any row for a per-model radar and full stat line. Blank cells (—) are figures we haven't verified.
✓ Verified data — every figure is cited. See sources.
| ▸ | 1 |
Claude Fable 5Anthropic
|
89.9 | 95 | 83 | 90 | $20 | 67 | 186s | 1M | Prop |
| ▸ | 2 |
Grok 4.5xAI
|
87.4 | — | — | 87 | $3 | 91 | 16.49s | 500K | Prop |
| ▸ | 3 |
Claude Opus 4.8Anthropic
|
85.5 | 89 | 79 | 88 | $10 | 60 | 23.13s | 1M | Prop |
| ▸ | 4 |
Claude Sonnet 5Anthropic
|
84.0 | 85 | — | 82 | $6 | 89 | 211s | 1M | Prop |
| ▸ | 5 |
GPT-5.5OpenAI
|
82.9 | 81 | 83 | 85 | — | — | — | — | Prop |
| ▸ | 6 |
MiniMax M3MiniMax
|
82.2 | — | — | 82 | $0.53 | 95 | 2.04s | 1M | Open |
| ▸ | 7 |
Kimi K2.7 CodeMoonshot AI
|
82.1 | — | — | 82 | $1.71 | 46 | 3.01s | 262K | Open |
| ▸ | 8 |
DeepSeek V4 ProDeepSeek
|
81.8 | 78 | — | 88 | $0.55 | 64 | 1.87s | 1M | Open |
| ▸ | 9 |
MiniMax M2.1MiniMax
|
81.8 | — | — | 82 | — | — | — | — | Prop |
| ▸ | 10 |
Qwen3.7 MaxAlibaba
|
81.5 | 77 | — | 87 | $3.75 | 204 | 2.55s | 1M | Prop |
| ▸ | 11 |
Gemini 3.5 FlashGoogle
|
80.9 | 79 | 76 | 88 | $3.38 | 167 | 25.21s | 1M | Prop |
| ▸ | 12 |
DeepSeek V3.2 (Thinking)DeepSeek
|
80.7 | — | — | 81 | — | — | — | — | Open |
| ▸ | 13 |
GPT-5.3 CodexOpenAI
|
79.6 | 75 | 78 | 87 | $4.81 | 87 | 92.81s | 400K | Prop |
| ▸ | 14 |
GLM 5.2Z.ai (Zhipu AI)
|
78.7 | 79 | — | — | — | — | — | — | Prop |
| ▸ | 15 |
Gemini 3.1 ProGoogle
|
78.0 | 76 | 71 | 89 | $4.5 | 147 | 32.59s | 1M | Prop |
| ▸ | 16 |
Claude Opus 4.7Anthropic
|
77.6 | 84 | 70 | — | — | — | — | — | Prop |
| ▸ | 17 |
Kimi K2.6Moonshot AI
|
76.7 | 77 | 67 | 87 | — | — | — | — | Open |
| ▸ | 18 |
Gemini 3 ProGoogle
|
73.5 | 73 | 74 | — | — | — | — | — | Prop |
| ▸ | 19 |
Grok 4.20xAI
|
57.3 | — | 57 | — | — | — | — | — | Prop |
| ▸ | 20 |
Claude Haiku 4.5Anthropic
|
49.7 | 67 | 36 | 41 | $2 | 96 | 11.91s | 200K | Prop |
| ▸ | 21 |
MiniMax M2.7MiniMax
|
45.1 | — | 45 | — | — | — | — | — | Prop |
| ▸ | 22 |
Qwen3-Coder 480B A35BAlibaba
|
— | — | — | — | — | — | — | — | Open |