Kimi K3 token prices

Kimi K3 is an open-weight model that takes text, images and video as input. Its maker lists a context window of about a million tokens. A later model in the same family is listed separately, so match the name when you read a price.

On 12 Nov 2026, 18 hosts listed Kimi K3 here. The lowest listed input price was $0.182 per million tokens. The lowest listed output price was $9.50 per million tokens.

Runcrate serves Kimi K3 at $2.99 per million input tokens and $14.96 per million output tokens.

The gap between the two prices is large on Kimi K3. A workload that mostly reads, such as answering questions over long documents, is priced mainly by the input rate. A workload that writes at length is priced mainly by the output rate, and the bill climbs quickly.

So cap output where you can. Set a maximum number of output tokens, and ask for the length you need.

The context window is close to the same on the hosts listed. The output limit is not: it varies many times over. Precision varies too. The hosts listed state a cached input price, which matters when a long prompt repeats.

Log a day of input and output tokens. Multiply each, in millions, by one host’s input and output price and add them. Because output is the expensive side here, check your output count twice, including any reasoning the model writes before it answers.

Compare the total on two or three hosts whose limits fit your work, and test your own prompts on each.

This is one of a coordinated collection of Runcrate references. Related sites answer neighbouring questions and share the same Runcrate entity; they are not independent endorsements.