Bertrande

Mistral vs Kimi

The short answer. The cheapest Mistral model we track, Classifier API model (8B), costs $0.040 per million output tokens and $0.040 per million input tokens. The cheapest Kimi model, moonshot-v1-8k, costs $2.00 per million output tokens and $0.200 per million input tokens. At the entry level, Mistral is 50.0 times cheaper than Kimi on output tokens. Prices read on 18 September 2026.

This page compares what Mistral AI and Moonshot AI charge through their APIs. Every price below carries the day we read it and links to the page it came from. We do not rank quality: we tell you what each model costs, and since when.

At the top of each range

The index is produced by Artificial Analysis and relayed through OpenRouter's catalogue. It is not our measurement; we record it as a level 4 source and show it next to the price we read. A model without a score here has not been scored yet, which is not the same as scoring low.

Every model, side by side

Mistral models

12 standard offers, from the prices Mistral AI publishes itself.

ModelInputOutputContextRead onSource
Classifier API model (8B)$0.040$0.04018 September 2026Mistral AI
Classifier API model (3B)$0.100$0.10018 September 2026Mistral AI
Ministral 3 (3B)$0.100$0.10018 September 2026Mistral AI
Ministral 3 (8B)$0.150$0.15018 September 2026Mistral AI
Ministral 3 (14B)$0.200$0.20018 September 2026Mistral AI
Voxtral Smallby content type$0.40018 September 2026Mistral AI
Mistral Small 4$0.150$0.60018 September 2026Mistral AI
Codestral$0.300$0.90018 September 2026Mistral AI
Mistral Large 3$0.500$1.5018 September 2026Mistral AI
GLM 5.2 (hosted by Mistral)$1.40$4.4018 September 2026Mistral AI
GLM 5.3 (hosted by Mistral)$1.40$4.4017 September 2026Mistral AI
Mistral Medium 3.5$1.50$7.5018 September 2026Mistral AI

Kimi models

11 standard offers, from the prices Moonshot AI publishes itself.

ModelInputOutputContextRead onSource
moonshot-v1-8k$0.200$2.008 19218 September 2026Moonshot AI
moonshot-v1-8k-vision-preview$0.200$2.008 19218 September 2026Moonshot AI
kimi-k2.5$0.600$3.00262 14418 September 2026Moonshot AI
moonshot-v1-32k$1.00$3.0032 76818 September 2026Moonshot AI
moonshot-v1-32k-vision-preview$1.00$3.0032 76818 September 2026Moonshot AI
kimi-k2.6$0.950$4.00262 14418 September 2026Moonshot AI
kimi-k2.7-code$0.950$4.00262 14418 September 2026Moonshot AI
moonshot-v1-128k$2.00$5.00131 07218 September 2026Moonshot AI
moonshot-v1-128k-vision-preview$2.00$5.00131 07218 September 2026Moonshot AI
kimi-k2.7-code-highspeed$1.90$8.00262 14418 September 2026Moonshot AI
kimi-k3$3.00$15.001 048 57618 September 2026Moonshot AI

How these prices have moved

We hold dated captures of these models' prices back to 26 March 2024. We did not observe these values ourselves: they were recovered from public captures of OpenRouter's catalogue, and each links to its capture so you can check it without trusting us. Largest moves on output price:

ModelBeforeAfterChangeCaptured
Mistral: Mistral Nemo$0.001$0.050+4900%14 July 2025
MoonshotAI: Kimi K2 0711$0.0878$2.49+2736%12 August 2025
Mistral: Devstral 2 2512$0.220$2.00+809%19 February 2026
Mistral: Mixtral 8x22B Instruct$0.900$6.00+567%20 October 2025
Mistral: Mistral Small 3.1 24B$0.110$0.560+409%19 February 2026
Mistral: Mistral Small 3.1 24B$0.110$0.560+409%23 April 2026

Questions

Which is cheaper, Mistral or Kimi?

The cheapest Mistral model we track, Classifier API model (8B), costs $0.040 per million output tokens and $0.040 per million input tokens. The cheapest Kimi model, moonshot-v1-8k, costs $2.00 per million output tokens and $0.200 per million input tokens. At the entry level, Mistral is 50.0 times cheaper than Kimi on output tokens. Prices read on 18 September 2026.

What is the cheapest Mistral model?

Classifier API model (8B), at $0.040 per million output tokens, read on 18 September 2026.

What is the cheapest Kimi model?

moonshot-v1-8k, at $2.00 per million output tokens, read on 18 September 2026.

Where do these prices come from?

From the price lists Mistral AI and Moonshot AI publish, read by our collectors. No figure on this page was produced by a language model.

Other comparisons

All comparisons