Model / inference API

Moonshot AI

Moonshot AI sells the Kimi API through its developer platform, now at platform.kimi.ai, and it suits teams who want Kimi K3 and K3.1 Pro with a 1,048,576-token context window straight from the maker.

Website (opens in a new tab)

Checked 2026-11-05

Moonshot AI sells the Kimi API through its developer platform, now at platform.kimi.ai, and it suits teams who want Kimi K3 and K3.1 Pro with a 1,048,576-token context window straight from the maker. The pricing page, as listed on 5 November 2026, bills per million tokens, but its price tables came through empty in the text we could read, so we give no rate. Hosts that resell Kimi K3 quote their own figures, for example $2.7 input and $13.5 output on SiliconFlow. The page says caching is automatic with a default five-minute lifetime, K3 cache writes are billed by 5-minute or 1-hour lifetime, and batch use on K3 models costs 60% of standard. K2 models, with 262,144 tokens of context, are to retire by the end of December 2026. The catch is the data term: the model use terms say content may be used to improve the services unless you agree otherwise in writing, and a restriction on training is offered through enterprise or separate written agreements. Retention periods and data location are not stated there. Figures come from Moonshot's pricing page and terms. We have not run it ourselves. We would use it to test Kimi K3 first-party before choosing a reseller.

Mentioned here

Questions

How much does Moonshot AI cost?

Pay as you go on the Kimi API, billed per token with separate input, output and cache rates and no subscription plan. Batch inference for kimi-k3.1-pro and kimi-k3 is billed at 60% of the standard price. The Kimi membership and Kimi Code are separate products (checked 5 November 2026).

Is Moonshot AI OpenAI compatible?

Yes, it offers an OpenAI-compatible API, so most agents and SDKs can point at it by changing the base URL.

Reviews

Loading reviews…

Write a review

Your rating

Up to 1000 characters. Stars alone are enough.

Similar

  • Featherless

    Featherless is a flat-rate subscription for open models, not a per-token meter, and it suits hobbyists and small teams who want to switch between many open models for one monthly price.

    • Model / inference API
    • OpenAI compatible

  • Z.AI

    Z.ai sells the GLM model family through its open platform API, and it suits developers who want GLM 5.3 straight from the maker, with free small models for testing.

    • Model / inference API
    • OpenAI compatible

  • Minimax

    MiniMax sells its model API through the MiniMax Open Platform, and it suits developers who want M3 or M3.1 Flash directly from the maker at a low token rate.

    • Model / inference API
    • OpenAI compatible

  • AI21

    AI21 sells the Jamba model family through an API, and it suits teams that want a long-context model from a vendor that says it does not train on customer inputs.

    • Model / inference API

  • Databricks

    Databricks sells model serving inside its data platform, so it suits teams that already keep data and pipelines on Databricks and want hosted open models next to them.

    • Model / inference API
    • OpenAI compatible

  • Nvidia

    NVIDIA's API platform at build.nvidia.com hosts models as NIM microservices on its DGX Cloud for trying out, and it suits developers who want to test Nemotron and other models before deciding where to run them.

    • Model / inference API