Kimi K2 Thinking pricing.

prices as of · 26 providers · 29 prices

Moonshot AI's Kimi K2 Thinking holds 262k tokens of context and has a reasoning mode. Open weights, so anyone may host it, which is why the prices below differ so much.

Kimi K2 Thinking costs $0.60 in and $2.50 out per million tokens from Moonshot AI; the cheapest price we can explain is $3.41 at Vercel AI Gateway, checked 16 Sept 2026.

tool callingreasoningstructured outputopen weights262k contextreleased 6 Nov 2025

Moonshot AI list price

$4.30/M tokens

$0.60 in · $2.50 out

read · 21% above the market median

Cheapest price we can explain

$3.41/M tokens

$0.47 in · $2 out · 21% under list

Vercel AI Gateway
View at Vercel AI Gateway

read

What it costs for your work

pages a month

Moonshot AI list price

$0.51a month

cheapest we can explain · Vercel AI Gateway

$0.402a month

difference

$0.108saved

A page is about 600 tokens in and 60 out, a support reply 200 in and 300 out, an email 150 in and 250 out. Nothing else is counted: no cache discount, no batch rate, no free tier.

Every provider selling Kimi K2 Thinking

Four groups, decided by the price and by what the provider publishes about it. Prices are per million tokens, and the share is that row against Moonshot AI's price.

The maker's list price

1 price

What the lab that trained the model charges on its own API.

  • Moonshot AI

    The price Moonshot AI charges for the model it trained. Every row below is measured against it.

    $0.60 / $2.50in / out per M

    1.00×of list price

Cheaper, and we know why

6 prices

Below list, with a published reason: hosting the open weights, a regional price list, credits, a stated discount.

  • Vercel AI Gateway

    Routes to a host serving the open weights; precision not disclosed.

    216k context hereVercel AI Gateway

    $0.47 / $2in / out per M

    0.79×of list price

  • Helicone

    Hosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.

    256k context here

    $0.48 / $2in / out per M

    0.80×of list price

  • Poe

    A consumer subscription with an API bolted on. Points convert to dollars, so the rate moves when the conversion does.

    216k context herePoe

    $0.47 / $2.02in / out per M

    0.80×of list price

  • OpenCode Zen

    Hosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.

    $0.40 / $2.50in / out per M

    0.86×of list price

  • Alibaba Cloud (China)

    4% under the list price, an ordinary reseller margin.

    $0.57 / $2.29in / out per M

    0.93×of list price

  • 302.AI

    4% under the list price, an ordinary reseller margin.

    $0.57 / $2.30in / out per M

    0.94×of list price

Same price as the maker

21 prices

The same price the maker charges, sold by someone else.

Cheaper, and nobody says why

1 price · never ranked

Below list on a closed model with nothing to account for it. Shown, never ranked, never a pick.

  • Featherless

    We could not read this provider's prices reliably, so the row is listed but never ranked.

    33k context hereFeatherless

    $0.60 / $2.50in / out per M

    1.00×of list price

Frequently asked

Who sells Kimi K2 Thinking cheapest?
Vercel AI Gateway, at $0.47 in and $2 out per million tokens, read on 16 Sept 2026. Routes to a host serving the open weights; precision not disclosed.
Is Kimi K2 Thinking open weights?
Yes. Moonshot AI publishes the weights, so anyone may host Kimi K2 Thinking, which is why 26 providers sell it and why their prices differ so much.
How much context does Kimi K2 Thinking hold?
262,144 tokens in one call, which is about 262k, and it can call your own functions.
How does Kimi K2 Thinking compare with the rest of the market?
At $4.30 per million tokens, Moonshot AI's price is 21% above the $3.55 median across the 337 text models we track.
More from Moonshot AIKimi K2.6Kimi K3Kimi K2.7 CodeKimi K2.5Similar contextKimi K2.6Kimi K2.7 CodeGLM-5.1Kimi K2.5