Kimi K3 pricing
What Kimi K3 costs per million tokens, on 3 platforms: input, output, cache read and cache write, with batch and long-context rates where the provider publishes them. Every figure is the rate Vigil's proxy bills with, read from the provider's own page.
Prices checked 9 Oct 2026 against platform.kimi.ai · registry sources: platform.kimi.ai, docs.fireworks.ai, baseten.co. All shown rates agreed with the page. The Moonshot row was read against Moonshot’s page. The Fireworks and Baseten rows carry the registry’s own retrieval dates and were not re-read on this date.
Kimi K3 on Moonshot API
Model id kimi-k3. Verified against the provider's published price; registry row retrieved 2026-09-11 from platform.kimi.ai.
| Applies | Input | Output | Cache read | Cache write |
|---|---|---|---|---|
| Standard | $3.00 | $15.00 | $0.300 | — |
USD per 1M tokens. A dash means the provider publishes no rate for that dimension, so the proxy declines to price it.
- Endpoint classes: global.
- on_demand: the standard rate.
- Above 200,000 input tokens the rate is unverified, so the proxy declines to price such a call rather than guess.
Kimi K3 on Fireworks AI
Model id fireworks/kimi-k3. Verified against the provider's published price; registry row retrieved 2026-09-11 from docs.fireworks.ai.
| Applies | Input | Output | Cache read | Cache write |
|---|---|---|---|---|
| Standard | $3.00 | $15.00 | $0.300 | — |
USD per 1M tokens. A dash means the provider publishes no rate for that dimension, so the proxy declines to price it.
- Endpoint classes: global.
- on_demand: the standard rate.
- batch: 50% below the standard rate.
- Above 200,000 input tokens the rate is unverified, so the proxy declines to price such a call rather than guess.
Kimi K3 on Baseten
Model id moonshotai/Kimi-K3. Verified against the provider's published price; registry row retrieved 2026-09-11 from baseten.co.
| Applies | Input | Output | Cache read | Cache write |
|---|---|---|---|---|
| Standard | $3.00 | $15.00 | $0.300 | — |
USD per 1M tokens. A dash means the provider publishes no rate for that dimension, so the proxy declines to price it.
- Endpoint classes: global.
- on_demand: the standard rate.
- Above 200,000 input tokens the rate is unverified, so the proxy declines to price such a call rather than guess.
Price a Kimi K3 call
Kimi K3 on Moonshot API, standard rate, priced now
Cache writes, batch, endpoint class and monthly volume: AI token calculator
Questions
- Is Kimi K3 the same price on Fireworks AI and Baseten as on Moonshot API?
- Yes. The input, output and cache-read rates the registry holds for Moonshot API, Fireworks AI, Baseten are the same per 1M tokens on the global endpoint class. A regional or multi-region endpoint, where one is published, carries its own factor (shown per platform below).
- What does a cache read cost on Kimi K3?
- A cache-read token on Moonshot API costs $0.300 per 1M tokens, against $3.00 for an ordinary input token. An ordinary input token costs 10 times as much. No cache-write rate is published for this model.
- Does Kimi K3 have a batch price?
- No batch rate is published for Kimi K3 on Moonshot API, so the proxy prices every call at the standard rate and does not guess at a discount.
Other models: Claude Opus 5 · Claude Sonnet 5 · Claude Haiku 4.5 · Claude Fable 5.1 · Claude Opus 4.8 · Claude Sonnet 4.6 · GPT-6 Sol · GPT-6 Luna · GPT-5.6 Sol · GPT-5.6 Terra · GPT-5.3 Codex · GPT-4o · GPT-4o mini · Gemini 3.1 Pro · Gemini 3.5 Flash · Gemini 3.8 Flash · DeepSeek V4 Pro · DeepSeek Flash · Grok 4.6