Moonshot AI · AI model API pricing
How much does Kimi K3 cost?
Moonshot's frontier model — priced identically to Claude Sonnet 5 ($3/$15), with a 1M-token context window.
Rate card
| Input tokens | $3 / 1M |
| Cached input | $0.30 / 1M |
| Output tokens | $15 / 1M |
| Context window | 1M |
What that means per month
| Workload | Tokens / month | Cost / month |
|---|---|---|
| Prototype | 1M in · 0.3M out | $7.50/mo |
| Small app | 10M in · 3M out | $75/mo |
| Production app | 50M in · 15M out | $375/mo |
| High volume | 200M in · 60M out | $1,500/mo |
| Heavy platform | 1B in · 300M out | $7,500/mo |
Standard on-demand rates from Moonshot AI's official pricing docs, last checked 2026-07-17. Cache-hit input bills at $0.30 per 1M. Monthly figures exclude caching and batch discounts. Confirm live pricing before committing.
Where Kimi K3 sits on price
At the reference month of 10M input and 3M output tokens, Kimi K3 costs $75, which makes it the 26th cheapest of the 30 models tracked here. That is 44× the price of Qwen-Flash, the cheapest model at this mix ($1.70). Among the 10 frontier-class models it ranks 6th, behind GLM-5.2 at $27.2.
Cache economics
Cached input on Kimi K3 bills at $0.30 per 1M — 10% of its own input rate, against a 10% median across the models here that publish a cached-read price. 2 models discount cached reads more steeply than this. On the reference month, serving 80% of that input from cache takes the input side of the bill from $30 to $8.40, leaving output untouched at $45 — which is why a high reuse rate changes the ranking above rather than just shaving the total.
Cache reads only. Writes are billed separately by some providers and are excluded here — see the rate note above for Moonshot AI's terms. Try it at your own hit rate on the AI cost calculator.
Cheaper alternatives
At a reference workload of 10M input / 3M output tokens a month, these cost less than Kimi K3 ($75/mo):
-
Gemini 3.1 Pro $56/mo -
GPT-5.6 Terra $56/mo -
Claude Sonnet 5 $50/mo