Long context
Reasoning that requires holding a large input in view at once. A large advertised context window is not evidence of this; these benchmarks measure whether the window is usable.
Ordered by how many of these benchmarks put a model in their top 10, then by its best placing. Each column keeps its own ranking; nothing is averaged.
| Model | In top 10 | Artificial Analysis LLM Leaderboard AA-LCR |
|---|---|---|
 Kimi K3 kimi | 1/1 | #1 88.7% |
 Step 5 stepfun | 1/1 | #2 88.3% |
 MiMo V2.6 Pro xiaomi | 1/1 | #3 86.3% |
 Claude Fable 5.1 anthropic | 1/1 | #4 85.3% |
| Claude Opus 5.5 anthropic | 1/1 | #5 84.7% |
 GPT 5.5 openai | 1/1 | #6 84.3% |
 DeepSeek V4.1 Flash deepseek | 1/1 | #7 84.0% |
| GPT 5.6 Sol openai | 1/1 | #8 84.0% |
| GPT 6.1 Sol openai | 1/1 | #9 84.0% |
 Gemini 3.8 Flash google | 1/1 | #10 84.0% |
A dash means the board does not list that model — not a score of zero. Entries not yet matched to a model are left out here; they still appear on their own board.