CAP
Long Context
League table — best at long context · 30
mean percentile across this capability's benchmarks| # | Model | League | Coverage | $/1M | tok/s |
|---|---|---|---|---|---|
| 1 | GPT-5.2 Codex | 100.0 | 1/1 | $4.81 | — |
| 2 | GPT-5 | 99.7 | 1/1 | $3.44 | — |
| 3 | GPT-5.1 | 99.4 | 1/1 | $3.44 | — |
| 4 | Kimi K3 | 99.1 | 1/1 | $6 | 33 |
| 5 | GPT-5.5 | 98.9 | 1/1 | $11.25 | — |
| 6 | Claude Opus 4.5 | 98.6 | 1/1 | $10 | — |
| 7 | GPT-5.4 | 98.3 | 1/1 | $5.63 | — |
| 8 | KAT-Coder-Pro V1 | 98.0 | 1/1 | — | — |
| 9 | MiniMax-M3 | 97.7 | 1/1 | $0.525 | 87 |
| 10 | GPT-5.3 Codex | 97.4 | 1/1 | $4.81 | 126 |
| 11 | GPT-5.6 Luna (max) | 97.1 | 1/1 | $2.25 | 171 |
| 12 | GPT-5.6 Terra (max) | 96.8 | 1/1 | $5.63 | 128 |
| 13 | GPT-5.6 Sol (max) | 96.6 | 1/1 | $11.25 | 74 |
| 14 | MiMo-V2.5-Pro | 96.3 | 1/1 | $0.544 | 65 |
| 15 | GPT-5.2 | 96.0 | 1/1 | $4.81 | — |
| 16 | Gemini 3.1 Pro Preview | 95.7 | 1/1 | $4.5 | 132 |
| 17 | GLM-5.2 (max) | 95.4 | 1/1 | $2.15 | 157 |
| 18 | GPT-5.6 Terra | 95.1 | 1/1 | $5.63 | 120 |
| 19 | GPT-5.6 Sol | 94.8 | 1/1 | $11.25 | 64 |
| 20 | Claude Opus 4.6 | 94.5 | 1/1 | $10 | — |
| 21 | Gemini 3 Pro Preview | 94.3 | 1/1 | $4.5 | — |
| 22 | Claude Sonnet 4.6 | 94.0 | 1/1 | $6 | — |
| 23 | Claude Sonnet 5 | 93.7 | 1/1 | $4 | 83 |
| 24 | Claude Opus 4.7 | 93.4 | 1/1 | $10 | — |
| 25 | Claude 4.5 Haiku | 93.1 | 1/1 | $2 | 150 |
| 26 | Claude Opus 5 | 92.8 | 1/1 | $10 | 44 |
| 27 | Claude Fable 5 | 92.5 | 1/1 | $20 | 58 |
| 28 | Qwen3.6 Max Preview | 92.2 | 1/1 | $2.92 | — |
| 29 | Qwen3.6 Plus | 92.0 | 1/1 | $1.13 | 53 |
| 30 | Kimi K2.6 | 91.7 | 1/1 | $1.71 | — |
League = a model's mean percentile across the 1 benchmarkin this capability (best score per benchmark; partial coverage shown). Price & speed via Artificial Analysis.
Benchmarks in this capability (1)
AA Long-Context Reasoning349 scores