Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models739Deals30d257Policy30d64, 14 pendingSources1,010

MiniMax M2.7

MiniMaxReleased Mar 18, 2026Current

Score basis

MiniMax M2.7 is capable in long context and coding and behind the leaders in instruction following and reasoning. Too few results yet to rate agentic tasks, safety, math, multimodal tasks, multilingual tasks, or factuality.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Capable2 capabilities
  1. Long Context−14.0%1 of 3
  2. Coding−23.4%8 of 10
Limited2 capabilities
  1. Instruction Following−29.0%2 of 3
  2. Reasoning−32.0%3 of 6
Not rated6 capabilities

Too few results yet to rate Agentic, Safety, Math, Multimodal, Multilingual or Factuality.

Price

$0.30input$1.20outputper million tokens

From Artificial Analysis · 3 providers tracked · All prices

Cheaper than 59% of 331 priced models · 3:1 input-to-output blend, log scale

Evidence

99results on79benchmarks

  • 5 independently verified
  • 20 aggregator
  • 8 vendor-reported
  • 66 cross-referenced

From 9 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingJSON modeReasoning

As listed by OpenRouter

Research

8 papers reference MiniMax M2.7

MiniMax M2.7 benchmark results

99 results on 79 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Long Context

CapableFull long context ranking

14.0% behind the leader1 of 3 ranked benchmarks measured

Coding

CapableFull coding ranking

23.4% behind the leader8 of 10 ranked benchmarks measured

Show 6 more coding resultsHide 6 coding results

Instruction Following

LimitedFull instruction following ranking

29.0% behind the leader2 of 3 ranked benchmarks measured

Show 1 more instruction following resultHide 1 instruction following result

Reasoning

LimitedFull reasoning ranking

32.0% behind the leader3 of 6 ranked benchmarks measured

Show 7 more reasoning resultsHide 7 reasoning results

Agentic

Not enough dataFull agentic ranking

0 of 7 ranked benchmarks measured

Show 6 more agentic resultsHide 6 agentic results

Math

Not enough dataFull math ranking

0 of 5 ranked benchmarks measured

Show 2 more math resultsHide 2 math results

Factuality

Not enough dataFull factuality ranking

0 of 4 ranked benchmarks measured

Show 1 more factuality resultHide 1 factuality result

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 47 more resultsHide 47 results

MiniMax M2.7: common questions

Who makes MiniMax M2.7?

MiniMax M2.7 is made by MiniMax.

When was MiniMax M2.7 released?

MiniMax M2.7 was released on Mar 18, 2026, according to Artificial Analysis.

What is MiniMax M2.7 good at?

MiniMax M2.7 is capable in long context and coding and behind the leaders in instruction following and reasoning. Too few results yet to rate agentic tasks, safety, math, multimodal tasks, multilingual tasks, or factuality.

How much does MiniMax M2.7 cost?

MiniMax M2.7 costs $0.30 per million input tokens and $1.20 per million output tokens, according to Artificial Analysis. We track its price at 3 providers. At a mix of three input tokens to one output token, it is cheaper than 59% of the 331 priced models we track.

How many benchmarks has MiniMax M2.7 been tested on?

We track 99 results for MiniMax M2.7 on 79 benchmarks from 9 sources, 5 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does MiniMax M2.7 support?

OpenRouter lists tool calling, json mode, and reasoning for MiniMax M2.7.

About this record

Where MiniMax M2.7's numbers come from, and every name it appears under.

Tracked since
Apr 27, 2026
Newest source mention
Sep 3, 2026

Where the results come from

Verification: 99 scores · 5 independently verified · 20 aggregator-attributed · 66 vendor cross-reference · 8 vendor-reported. How these tiers are assigned

From 9 sources on 6 sites. Hugging Face supplies 53 of them; the 5 independently verified results come from 2 sites. Bars are coloured by trust tier.

  • huggingface.co53
  • artificialanalysis.ai20
  • thinkingmachines.ai13
  • api.llm-stats.com8
  • raw.githubusercontent.com4
  • datasets-server.huggingface.co1

Also known as

minimax-m2p7minimaxai/minimax-m2.7minimax-m2-7minimax-m2.7:freeMiniMax 2.7 230B A10Bminimax-m2.7-turbo