Kimi K2.6
Moonshot AI released 2026-04-20
MiniMax M3 is both better and cheaper.
It scores +0.3 higher and costs 46% less ($0.525/M against $0.976/M) on this workload — and it does everything this model does.
DeepSeek V4 Flash 0731 is cheaper still (89% less) but drops no image input.
Every price dimension
| Input | $0.541/M |
|---|---|
| Output | $2.28/M |
| Cached input | $0.091/M |
Independent scores
| Intelligence | 45.1 |
|---|---|
| Coding | 61.8 |
| Agentic | 31.2 |
| LMArena Elo | 1455default |
Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.
Best rankings by task
- agentichtmlslides #2of 16 1248
- agenticslides(html) #2of 16 1252
- agenticslides #5of 16 1187
- agenticslides(python-pptx) #5of 16 1186
- pptxslides #5of 14 1181
- webapps #8of 56 1268
- htmlslides #9of 34 1201
- 3d #13of 136 1315
Capability
| Context window | 262K |
|---|---|
| Max output | 262K |
| Input modes | text, image |
| Tool use | yes |
| Reasoning | optional |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-08-24 |
| Quality data | verified |
Kimi K2.6 is Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. It handles complex end-to-end coding tasks across Python, Rust, and Go, and...