Moonshot AI API pricing
Moonshot AI sells API access to the Kimi model family. Prices are per 1M tokens; the open-weight Kimi models are also sold by third-party hosts at their own prices.
Official price list: platform.moonshot.ai
Current lineup
How Moonshot AI billing works
Kimi models are open-weight, so the same model is often available from Fireworks, Together, and others — each at its own price. Compare the host rows on the model pages.
Every Moonshot AI model we price
The full catalog — including dated snapshots and variants — exactly as Marginal prices incoming usage events.
| Cached input / 1M | ||||
|---|---|---|---|---|
| kimi-k2-0711-preview | $0.60 | $2.50 | 128K | $0.15 |
| kimi-k2-0905-preview | $0.60 | $2.50 | 256K | $0.15 |
| kimi-k2-thinking | $0.60 | $2.50 | 256K | $0.15 |
| kimi-k2-thinking-turbo | $1.15 | $8.00 | 256K | $0.15 |
| kimi-k2-turbo-preview | $1.15 | $8.00 | 256K | $0.15 |
| kimi-k2.5 | $0.60 | $3.00 | 256K | $0.10 |
| kimi-k2.6 | $0.95 | $4.00 | 256K | $0.16 |
| kimi-latest | $2.00 | $5.00 | 128K | $0.15 |
| kimi-latest-128k | $2.00 | $5.00 | 128K | $0.15 |
| kimi-latest-32k | $1.00 | $3.00 | 32K | $0.15 |
| kimi-latest-8k | $0.20 | $2.00 | 8K | $0.15 |
| kimi-thinking-preview | $0.60 | $2.50 | 128K | $0.15 |
| moonshot-v1-128k | $2.00 | $5.00 | 128K | — |
| moonshot-v1-128k-0430 | $2.00 | $5.00 | 128K | — |
| moonshot-v1-128k-vision-preview | $2.00 | $5.00 | 128K | — |
| moonshot-v1-32k | $1.00 | $3.00 | 32K | — |
| moonshot-v1-32k-0430 | $1.00 | $3.00 | 32K | — |
| moonshot-v1-32k-vision-preview | $1.00 | $3.00 | 32K | — |
| moonshot-v1-8k | $0.20 | $2.00 | 8K | — |
| moonshot-v1-8k-0430 | $0.20 | $2.00 | 8K | — |
| moonshot-v1-8k-vision-preview | $0.20 | $2.00 | 8K | — |
| moonshot-v1-auto | $2.00 | $5.00 | 128K | — |
Prices last synced Aug 20, 2026 from the LiteLLM community price catalog — the same catalog Marginal prices real usage events against, refreshed daily. Spot an error? Tell us.
Other providers
Using Moonshot AI in production?
Marginal prices every call your app makes against this catalog and slices spend by customer, feature, or any field you define. One track() call to integrate.