Kimi K3 API pricing — cost per million tokens
Moonshot's Kimi K3 costs $3 per million input tokens and $15 per million output tokens on the standard API tier (verified 2026-07-23). Note: 2.8T open-weight MoE; always reasons, so reasoning traces bill at the output rate. Cache-hit input is $0.30. Compare it in the full pricing table or price your own prompts in the token cost calculator.
Kimi K3 price per million tokens
| direction | $ / Mtok | $ / 1K tokens |
|---|---|---|
| input (prompt) | $3 | $0.003 |
| output (completion) | $15 | $0.015 |
What real requests cost on Kimi K3
| workload | input | output | total |
|---|---|---|---|
| Chat message (1K in / 500 out) | $0.003 | $0.0075 | $0.0105 |
| Long document (100K in / 2K out) | $0.30 | $0.03 | $0.33 |
| Agent session (500K in / 50K out) | $1.50 | $0.75 | $2.25 |
How Kimi K3 pricing compares
On the same chat-message workload, Kimi K3 costs 37× more than DeepSeek V4 Flash (the cheapest tracked model) — capability, latency and context limits are the other side of that trade. Output tokens dominate long generations: at $15/Mtok, a 50K-token agent transcript costs $0.75 in output alone. Official rate card: Moonshot pricing.
Related: DeepSeek V4 Flash pricing · GPT-5.4 nano pricing · Gemini 2.5 Flash pricing · release timeline
The weekly e/acc newsletter
One email a week: what accelerated.
Frontier releases, price drops, compute buildouts — the week's acceleration in five minutes, sourced and numeric. Free.