API pricing
Kimi API billing based on token consumption, with model-specific and feature-specific pricing.
Billing basics
- Per-token billing: Each API call is billed separately for input tokens and output tokens.
- Token unit: 1M = 1,000,000 tokens.
- Model-specific pricing: Higher-capability models have higher per-token costs — choose the model that best fits your use case.
Additional feature billing
| Feature | Additional Fee |
|---|---|
| Web Search | $0.004 per invocation (independent of token consumption) |
Context caching
Context Caching allows you to cache frequently used context content (such as system prompts and reference documents). Tokens that hit the cache are billed at a discounted rate, effectively reducing costs for repetitive context.
Refer to the official documentation for detailed Context Caching pricing.
How is Kimi K3 billed?
Kimi K3 offers a 1M-token context and uses flat pay-as-you-go pricing — there is no tiering by context length. Input (with separate rates for cache hits and misses) and output are billed at uniform per-token prices. See Kimi K3 pricing.
Pricing details
For the complete model pricing table and billing rules:
You can go to platform.kimi.ai/docs/pricing/chat
Cost optimization tips
- Set the
max_tokensparameter appropriately to avoid unnecessarily long outputs. - Use Context Caching for repetitive system prompts and context.
- Choose the right model for the task complexity — use lightweight models for simple tasks.
- Streamline your prompt design to minimize unnecessary input tokens.
- If cost is a concern, you can use the
kimi-k2.6model.