Official site ◆ · 🇨🇳 China
| Tier | $/month | Included quota |
|---|---|---|
| Moderato | $None | — |
| Allegretto | $None | — |
| Allegro | $None | — |
| Vivace | $None | — |
readable by API partly convertible
As advertised. ~2048 Kimi Code requests/week (Moderato) scaling with credit multipliers 1x / 5x / 15x / 30x -> ~2,048 / ~10,240 / ~30,720 / ~61,440 per week
What that actually means. Progress since last check: the plans DO have a numeric quota after all -- ~2,048 Kimi Code requests/week on Moderato, scaling by the published 5x/15x/30x credit multipliers. But it is still not tokens. A 'request' is one agent turn and can carry anywhere from a few thousand to ~1M+ tokens (context up to 256K/1M plus output up to 1M), and no tokens-per-request or credit-to-token rate is published, so the plan cannot be converted to a token quota -- same limitation as Alibaba's request-metered Coding Plan. The live /coding/v1/usages endpoint (api.kimi.com) returns your limit but requires an active subscription; our key returned permission_denied, so it could not be read directly. Moonshot has announced a restructure separating Kimi and Kimi Code benefits, so these may change.
Where to check what is left. GET /coding/v1/usages (api.kimi.com; needs active sub)
checked 2026-08-13 · medium (prices) / low (quota): request counts and multipliers are from third-party write-ups, not Kimi's official page, which still publishes only adjectives · source 1 source 2 source 3
| Model | Coding index | tok/s | $/M in | $/M out | $ / session | Cache |
|---|---|---|---|---|---|---|
| Kimi K3 | 76.2 | 36.5 | $3.00 | $15.00 | $2.0958 | 92.0% |
| Kimi K2.6 | 61.8 | 0.0 | $0.56 | $2.36 | $0.8180 | 91.2% |
| Kimi K2.7 Code | 60.8 | 36.8 | $0.71 | $3.50 | $1.0948 | 94.2% |
| Kimi K2.5 | 46.8 | 0.0 | $0.57 | $2.85 | — | — |
5 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| Kimi K2.5 | int4 | $0.60 | $3.00 | 262,144 | 100.0% |
| Kimi K2.6 | int4 | $0.95 | $4.00 | 262,144 | 100.0% |
| Kimi K2.7 Code | int4 | $0.95 | $4.00 | 262,144 | 100.0% |
| Kimi K2.7 Code | int4 | $1.90 | $8.00 | 262,144 | 100.0% |
| Kimi K3 | mxfp4 | $3.00 | $15.00 | 1,048,576 | 100.0% |
Everyone serving these models, cheapest first, with the numeric precision each one runs. The lab is often not the cheapest place to buy its own model.
| Model | Served by | Quant | $/M in | $/M out | Uptime 30m |
|---|---|---|---|---|---|
| Kimi K3 | Morph | fp4 | $2.80 | $14.00 | 99.9% |
| Kimi K3 | DeepInfra | bf16 | $2.85 | $14.25 | — |
| Kimi K3 | DigitalOcean | unknown | $2.85 | $14.25 | 99.8% |
| Kimi K3 | BaseTen | fp8 | $3.00 | $15.00 | 99.0% |
| Kimi K3 | Chutes | mxfp4 | $3.00 | $15.00 | 98.1% |
| Kimi K3 | Fireworks | unknown | $3.00 | $15.00 | 99.6% |
| Kimi K3 | Modal | mxfp4 | $3.00 | $15.00 | 98.9% |
| Kimi K3 | Moonshot | mxfp4 | $3.00 | $15.00 | 100.0% |
| Kimi K3 | Phala | unknown | $3.00 | $15.00 | 99.4% |
| Kimi K3 | Sail Research | fp4 | $3.00 | $15.00 | 99.6% |
| Kimi K3 | Together | unknown | $3.00 | $15.00 | 99.1% |
| Kimi K3 | Wafer | unknown | $3.00 | $15.00 | 98.4% |
| Kimi K3 | Fireworks | unknown | $4.50 | $22.50 | 99.6% |
| Kimi K3 | Morph | fp4 | $6.00 | $22.50 | — |
| Kimi K2.6 | Baidu | fp4 | $0.56 | $2.36 | 100.0% |
| Kimi K2.6 | Decart | fp4 | $0.57 | $2.39 | 99.6% |
| Kimi K2.6 | StreamLake | fp8 | $0.60 | $2.52 | 98.5% |
| Kimi K2.6 | Chutes | int4 | $0.58 | $3.40 | 99.6% |
| Kimi K2.6 | Inceptron | int4 | $0.60 | $3.41 | 100.0% |
| Kimi K2.6 | CoreWeave | fp4 | $0.65 | $3.41 | 99.9% |
| Kimi K2.6 | DigitalOcean | unknown | $0.76 | $3.20 | 99.6% |
| Kimi K2.6 | Crusoe | bf16 | $0.70 | $3.50 | 88.2% |
| Kimi K2.6 | SiliconFlow | fp8 | $0.77 | $3.40 | 99.7% |
| Kimi K2.6 | DeepInfra | fp4 | $0.75 | $3.50 | 100.0% |
| Kimi K2.6 | Parasail | int4 | $0.75 | $3.50 | 99.6% |
| Kimi K2.6 | Venice | int4 | $0.75 | $3.50 | 81.4% |
| Kimi K2.6 | Novita | unknown | $0.80 | $3.40 | 99.9% |
| Kimi K2.6 | AtlasCloud | int4 | $0.95 | $4.00 | 91.1% |
| Kimi K2.6 | BaseTen | fp4 | $0.95 | $4.00 | — |
| Kimi K2.6 | Cloudflare | unknown | $0.95 | $4.00 | 100.0% |
| Kimi K2.6 | Fireworks | unknown | $0.95 | $4.00 | — |
| Kimi K2.6 | Moonshot | int4 | $0.95 | $4.00 | 100.0% |
| Kimi K2.6 | Sail Research | int4 | $1.00 | $4.00 | 100.0% |
| Kimi K2.6 | Phala | unknown | $1.09 | $4.60 | 96.3% |
| Kimi K2.6 | Together | unknown | $1.20 | $4.50 | 95.0% |
| Kimi K2.7 Code | Inceptron | int4 | $0.67 | $3.40 | 99.9% |
| Kimi K2.7 Code | DeepInfra | fp4 | $0.68 | $3.40 | 100.0% |
| Kimi K2.7 Code | Ambient | unknown | $0.69 | $3.49 | 99.7% |
| Kimi K2.7 Code | CoreWeave | int4 | $0.71 | $3.50 | 100.0% |
| Kimi K2.7 Code | Venice | int4 | $0.75 | $3.50 | 100.0% |
| Kimi K2.7 Code | Parasail | int4 | $0.76 | $3.50 | 99.8% |
| Kimi K2.7 Code | ModelRun | fp4 | $0.85 | $3.75 | 99.7% |
| Kimi K2.7 Code | SiliconFlow | fp8 | $0.86 | $3.80 | 100.0% |
| Kimi K2.7 Code | Novita | int4 | $0.91 | $3.84 | 100.0% |
| Kimi K2.7 Code | Alibaba / Qwen | fp8 | $0.95 | $4.00 | 100.0% |
| Kimi K2.7 Code | AtlasCloud | int4 | $0.95 | $4.00 | — |
| Kimi K2.7 Code | Cloudflare | unknown | $0.95 | $4.00 | — |
| Kimi K2.7 Code | Moonshot | int4 | $0.95 | $4.00 | 100.0% |
| Kimi K2.7 Code | Together | unknown | $0.95 | $4.00 | 98.7% |
| Kimi K2.7 Code | Moonshot | int4 | $1.90 | $8.00 | 100.0% |
| Kimi K2.5 | DigitalOcean | unknown | $0.38 | $2.02 | 98.4% |
| Kimi K2.5 | DeepInfra | fp4 | $0.45 | $2.25 | 100.0% |
| Kimi K2.5 | SiliconFlow | int4 | $0.45 | $2.25 | 100.0% |
| Kimi K2.5 | AtlasCloud | int4 | $0.49 | $2.50 | 99.8% |
| Kimi K2.5 | StreamLake | fp8 | $0.54 | $2.70 | 99.9% |
| Kimi K2.5 | Novita | unknown | $0.57 | $2.85 | 97.5% |
| Kimi K2.5 | Amazon Bedrock | unknown | $0.60 | $3.00 | 99.6% |
| Kimi K2.5 | Moonshot | int4 | $0.60 | $3.00 | 100.0% |
| Kimi K2.5 | Phala | unknown | $0.60 | $3.00 | — |
| Kimi K2.5 | Venice | unknown | $0.56 | $3.50 | 100.0% |