MMoonshot AI · Released Nov 6, 2025
Kimi K2 Thinking
Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...
- Released
- Nov 6 2025
- Context window
- 262K 98K out
- Input / 1M tokens
- $0.60 from $0.40
- Output / 1M tokens
- $2.50 from $2.00
Where to run it
19 providers compared
Cheapest: Helicone at $0.48 / $2.00 per 1M input / output tokens — 42% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| Helicone | $0.48 | $2.00 | — | 256K | 262K | — | Cheapest |
| OpenCode | $0.40 | $2.50 | $0.40 | 262K | 262K | — | +8% |
| Alibaba (China) | $0.57 | $2.29 | — | 262K | 16K | — | +17% |
| 302.AI | $0.57 | $2.30 | — | 262K | 262K | — | +17% |
| Qiniu | $0.58 | $2.32 | — | 256K | 100K | — | +18% |
| NovitaBF16 | $0.60 | $2.50 | $0.15 | 262K | 98K | 100%24h | +25% |
| $0.60 | $2.50 | — | 262K | 236K | 90.2%24h | +25% | |
| NanoGPT | $0.60 | $2.50 | $0.15 | 262K | 98K | — | +25% |
| Kilo Gateway | $0.60 | $2.50 | $0.15 | 262K | 98K | — | +25% |
| ZenMux | $0.60 | $2.50 | $0.15 | 262K | 64K | — | +25% |
| IO.NET | $0.60 | $2.50 | $0.30 | 33K | 4K | — | +25% |
| Charm Hyper | $0.60 | $2.50 | $0.30 | 262K | 26K | — | +25% |
| Novita AI | $0.60 | $2.50 | $0.15 | 262K | 98K | — | +25% |
| Amazon Bedrock | $0.60 | $2.50 | — | 262K | 16K | — | +25% |
| Hugging Face | $0.60 | $2.50 | $0.15 | 262K | 262K | — | +25% |
| Merge Gateway | $0.60 | $2.50 | — | 262K | 262K | — | +25% |
| Eden AI | $0.60 | $2.50 | — | 256K | 262K | — | +25% |
| Meganova | $0.60 | $2.60 | — | 262K | 262K | — | +28% |
| AI/ML API | $0.82 | $3.44 | $0.21 | 262K | 98K | — | +72% |
Value
Price vs. performance
Kimi K2 Thinking Other Moonshot AI Other labs Value frontier · Terminal-Bench vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
2 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| Terminal-BenchSoftware engineering · Tasks completed with the programs in a real terminal (best agent) | 35.7% | #27of 40 | GPT-5.584.7% |
| METR time horizonSoftware engineering · Length of software task solved more often than not | 54 min | #13of 23 | Claude Opus 4.612 h |
Specs
Details
- Model ID
moonshotai/kimi-k2-thinking- Developer
- Moonshot AI
- Release date
- Nov 6, 2025 ·
- Context window
- 262,144 tokens
- Max output
- 98,304 tokens
- API price
- $0.60 / $2.50 per 1M input / output tokens
- Providers
- 19
- Input types
- text
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.