Z.ai

Kimi K3

Kimi K3 is a Kimi model from Z.ai, released in 2026-07. It costs $3 / $15 per 1M, has a 1.0M-token context window, and is best for coding, frontend, agents. Last verified 2026-07-18.

Try Kimi K3 Docs Take the picker quiz

Spec sheet

Pricing

Input
$3 / 1M
Output
$15 / 1M
Cached input
$0.3 / 1M
Free tier
No

Context & speed

Context window
1.0M tokens
Max output
131k tokens
Throughput
~85 tok/s
Time to first token
~700 ms
Speed tier
balanced

Capabilities

Tool use
Yes
Structured output
Yes
Prompt caching
Yes
Extended thinking
Yes
Vision input
Yes
Audio in / out
No
Fine-tuning
No

Deployment

Open weights
No
On-prem
No
HIPAA eligible
No
Zero retention
No
Regions
apac, us

Estimated monthly cost

Assumes typical token shape: 2k input, 600 output per call. Prompt caching is excluded from these figures.

10k calls/mo
$150.00
per month
100k calls/mo
$1.5k
per month
1M calls/mo
$15k
per month

When to use Kimi K3

Sweet spot

  • coding
  • frontend
  • agents
  • complex reasoning
  • long context
  • multimodal

Known trade-offs

  • data-routing via China for hosted API
  • weights not yet released (promised 2026-07-27)
  • priciest model from a Chinese lab to date

Works with

Moonshot SDKOpenAI-compatible APIOpenRouterLangChain

Compare Kimi K3 to other models

vs Claude Sonnet 5vs GPT-5.6 Terravs Gemini 3.1 Provs Gemini 3.5 Flashvs Qwen3.7-Maxvs GPT-5.6 Luna

FAQ — Kimi K3

How much does Kimi K3 cost?

Kimi K3 costs $3 / $15 per 1M tokens on the Z.ai API. Cached input reads cost $0.3 per 1M, cutting the input bill by roughly 90% on repeat system prompts.

What is the context window of Kimi K3?

Kimi K3 has a 1.0M-token context window with up to 131k tokens of output. That's enough for entire codebases, long transcripts, or multi-document RAG.

Does Kimi K3 have a free tier?

No — No free API tier; kimi.com web chat is free with limits.

Is Kimi K3 HIPAA / EU / on-prem friendly?

Kimi K3 is not HIPAA-eligible, not available in an EU region, and is API-only. Zero data retention is not available.

What is Kimi K3 best for?

Kimi K3 is best for coding, frontend, agents, complex reasoning. Trade-offs to be aware of: data-routing via China for hosted API; weights not yet released (promised 2026-07-27); priciest model from a Chinese lab to date.

Which tools and SDKs work with Kimi K3?

Kimi K3 integrates with Moonshot SDK, OpenAI-compatible API, OpenRouter, LangChain. Most major AI frameworks support it either natively or through OpenAI-compatible endpoints.