MiniMax-M2.5 pricing.

prices as of · 48 providers · 59 prices

MiniMax-M2.5 holds 205k tokens of context and has a reasoning mode. Open weights, so anyone may host it, which is why the prices below differ so much.

MiniMax-M2.5 costs $0.30 in and $1.20 out per million tokens from MiniMax; the cheapest price we can explain is $1.54 at DInference, checked 16 Sept 2026.

tool callingreasoningopen weights205k contextreleased 12 Feb 2026

MiniMax list price

$2.10/M tokens

$0.30 in · $1.20 out

read · 41% below the market median

Cheapest price we can explain

$1.54/M tokens

$0.22 in · $0.88 out · 27% under list

DInference

read

What it costs for your work

pages a month

MiniMax list price

$0.252a month

cheapest we can explain · DInference

$0.185a month

difference

$0.0672saved

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 MiniMax-M2.5

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 MiniMax's price.

The maker's list price

1 price

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

  • MiniMax

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

    $0.30 / $1.20in / out per M

    1.00×of list price

Cheaper, and we know why

10 prices

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

  • DInference

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

    200k context here

    $0.22 / $0.88in / out per M

    0.73×of list price

  • DeepInfra

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

    197k context hereDeepInfra

    $0.15 / $1.15in / out per M

    0.76×of list price

  • TrustedRouter

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

    197k context hereTrustedRouter

    $0.23 / $0.95in / out per M

    0.78×of list price

  • GreenPT

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

    $0.19 / $1.13in / out per M

    0.81×of list price

  • Venice

    10% under the list price, an ordinary reseller margin.

    198k context hereVenice

    $0.27 / $0.95in / out per M

    0.84×of list price

  • Nous Portal

    10% under the list price, an ordinary reseller margin.

    $0.27 / $0.95in / out per M

    0.84×of list price

  • Poe

    9% under the list price, an ordinary reseller margin.

    197k context herePoe

    $0.27 / $0.96in / out per M

    0.85×of list price

  • OpenRouter

    10% under the list price, an ordinary reseller margin.

    $0.27 / $1.08in / out per M

    0.90×of list price

  • StreamLake

    10% under the list price, an ordinary reseller margin.

    200k context here

    $0.27 / $1.08in / out per M

    0.90×of list price

  • Avian

    10% under the list price, an ordinary reseller margin.

    197k context hereAvian

    $0.27 / $1.08in / out per M

    0.90×of list price

Same price as the maker

38 prices

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

showing the 25 cheapest of 38 · every row in the open dataset

Cheaper, and nobody says why

2 prices · 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.

    $0.29 / $1.20in / out per M

    0.99×of list price

  • FriendliAI

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

    197k context hereFriendliAI

    $0.30 / $1.20in / out per M

    1.00×of list price

Given away with limits

8 free tiers · never ranked

These providers give MiniMax-M2.5 away inside a quota of their own. A limit is not a price, so none of them counts as cheapest here or anywhere else on the site.

  • Alibaba Token Plan
  • Alibaba Coding Plan (China)
  • MiniMax Token Plan (minimaxi.com)
  • MiniMax Token Plan (minimax.io)
  • Alibaba Coding Plan
  • SCNet Token Plan
  • Tencent Coding Plan (China)
  • Alibaba Token Plan (China)

Frequently asked

Who sells MiniMax-M2.5 cheapest?
DInference, at $0.22 in and $0.88 out per million tokens, read on 16 Sept 2026. Hosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.
Is MiniMax-M2.5 open weights?
Yes. MiniMax publishes the weights, so anyone may host MiniMax-M2.5, which is why 48 providers sell it and why their prices differ so much.
How much context does MiniMax-M2.5 hold?
204,800 tokens in one call, which is about 205k, and it can call your own functions.
How does MiniMax-M2.5 compare with the rest of the market?
At $2.10 per million tokens, MiniMax's price is 41% below the $3.55 median across the 337 text models we track.
More from MiniMaxMiniMax-M3MiniMax-M2.7MiniMax-M2.1MiniMax-M2Similar contextKimi K2.6Kimi K2.7 CodeGLM-5.1DeepSeek V3.2