# Moonshot has stopped offering its Kimi K2 models to new API users and plans to retire them all by the end of December.

> A Kimi API notice says K2 models are closed to new users from Oct 31, and Moonshot plans to retire all K2 models by the end of December.

Published 2026-11-01T12:25:00-08:00 · News · Canonical: https://news.littlewolflabs.com/posts/2026-11-01-kimi-k2-retirement

A notice on Moonshot's Kimi API platform says that starting October 31, 2026, "K2 series models will no longer be available to newly registered users; existing users are not affected," and that it plans "to retire all K2 series models by the end of December 2026" ([Kimi pricing docs](https://platform.kimi.ai/docs/pricing/chat), [Kimi models list](https://platform.kimi.ai/docs/models)). No exact shutdown date is given. The K2 models still listed are kimi-k2.6, kimi-k2.7-code and the faster kimi-k2.7-code-highspeed, all with a 256K-token context window, meaning how much text the model can work with at once. The notice links to its suggested replacement, Kimi K3.1 Pro, which has a 1-million-token context window and a Highspeed version that Moonshot says runs 2 to 3 times faster ([K3.1 Pro quickstart](https://platform.kimi.ai/docs/guide/kimi-k3-1-pro-quickstart)). It doesn't drop in exactly like K2. Thinking mode, where the model reasons before it answers, is always on. You can lower how much it reasons with a `reasoning_effort` setting of low, high or max, and max is the default. Settings like temperature and top_p are fixed, so leave them out of your requests. For images, it doesn't accept public image URLs, so you send the image itself (base64) or upload it first. The page also says K3.1 Pro unlocks only after you add at least $1 to your account.

*When you'll feel it:* if you signed up before Oct 31, your K2 calls keep working for now, but you have until about the end of December to move. Test K3.1 Pro on your real prompts first, since always-on thinking can change speed and cost. Try `reasoning_effort` set to low if responses feel slow.