6 models tracked · known for Kimi K2.6, Kimi K2 (Jul 2025), Kimi K2 Thinking
Claims drawn from cited facts, not live model generation.
Moonshot AI is the lab covered by this summary.
1 cited fact
Moonshot AI's tracked footprint is six shipped models, signaling a model-focused lab. With only models tracked, the graph presents it as product-first rather than research-communication-heavy.
2 cited facts
69 benchmark scores across this lab’s models, tagged by evidence status.
| Model | Best score | Claim status |
|---|---|---|
| Kimi K3 | 97.22 | independently reproduced |
| Kimi K2.7 Code | 95.55 | independently reproduced |
| Kimi K2.5 | 92.19 | independently reproduced |
| Kimi K2 (Jul 2025) | 85.6 | unverified |
| Kimi K2 Thinking | 83.04 | independently reproduced |
Where a model from Moonshot AI holds the top score, with the benchmark’s integrity grade.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks