← Models
Kimi K2.6
Moonshot AI · open weight · kimi-k2.6
Kimi K2.6 is Moonshot AI's open-source, native multimodal agentic model focused on state-of-the-art coding, long-horizon execution, and agent swarm capabilities. It scales horizontally to 300 sub-agents executing 4,000 coordinated steps, dynamically decomposing tasks into parallel, domain-specialized subtasks. K2.6 unifies text, image, and video input with thinking and non-thinking modes, supports a 256K context, and powers proactive 24/7 background agents that manage schedules, execute code, and orchestrate cross-platform operations without human oversight.
Benchmark scores
| Benchmark | Score |
|---|---|
| V* | 0.97 |
| AIME 2026 | 0.96 |
| MathVision | 0.93 |
| HMMT Feb 26 | 0.93 |
| GPQA | 0.91 |
| CharXiv-R | 0.87 |
| BrowseComp | 0.86 |
| DeepSearchQA | 0.83 |
| Claw-Eval | 0.81 |
| WideSearch | 0.81 |
| SWE-Bench Verified | 0.80 |
| MMMU-Pro | 0.80 |
| SWE-bench Multilingual | 0.77 |
| OSWorld-Verified | 0.73 |
| Humanity's Last Exam | 0.36 |
Pricing
- Moonshot AI$0.95 / $4.00
- DeepInfra$0.75 / $3.50
- Novita$0.95 / $4.00
- Fireworks$0.95 / $4.00
Input / output per 1M tokens
AA metrics
No Artificial Analysis link yet.
Arena Elo
- code1509
