Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d271Policy30d66, 14 pendingSources1,015

MiniMax M2.1

MiniMaxReleased Dec 23, 2025Deprecated

Score basis

MiniMax M2.1 is behind the leaders in long context, instruction following, factuality, agentic tasks, reasoning, and coding. Too few results yet to rate safety, math, multimodal tasks, or multilingual tasks.

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.

Limited6 capabilities
  1. Long Context−27.5%1 of 3
  2. Instruction Following−28.9%1 of 3
  3. Factuality−36.5%3 of 4
  4. Agentic−37.2%2 of 7
  5. Reasoning−38.5%4 of 6
  6. Coding−38.8%6 of 10
Not rated4 capabilities

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

Price

$0.30input$1.20outputper million tokens

From Artificial Analysis · 3 providers tracked · All prices

Cheaper than 59% of 330 priced models · 3:1 input-to-output blend, log scale

Evidence

71results on53benchmarks

  • 7 independently verified
  • 15 aggregator
  • 35 vendor-reported
  • 14 cross-referenced

From 11 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingJSON modeReasoning

As listed by OpenRouter

Research

1 paper reference MiniMax M2.1

MiniMax M2.1 benchmark results

71 results on 53 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

LimitedFull long context ranking

27.5% behind the leader1 of 3 ranked benchmarks measured

Show 1 more long context resultHide 1 long context result

Instruction Following

LimitedFull instruction following ranking

28.9% behind the leader1 of 3 ranked benchmarks measured

Show 1 more instruction following resultHide 1 instruction following result

Factuality

LimitedFull factuality ranking

36.5% behind the leader3 of 4 ranked benchmarks measured

Agentic

LimitedFull agentic ranking

37.2% behind the leader2 of 7 ranked benchmarks measured

Show 2 more agentic resultsHide 2 agentic results

Reasoning

LimitedFull reasoning ranking

38.5% behind the leader4 of 6 ranked benchmarks measured

Show 4 more reasoning resultsHide 4 reasoning results

Coding

LimitedFull coding ranking

38.8% behind the leader6 of 10 ranked benchmarks measured

Show 5 more coding resultsHide 5 coding results

Math

Not enough dataFull math ranking

0 of 5 ranked benchmarks measured

More results

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

Show 34 more resultsHide 34 results

MiniMax M2.1: common questions

Who makes MiniMax M2.1?

MiniMax M2.1 is made by MiniMax.

When was MiniMax M2.1 released?

MiniMax M2.1 was released on Dec 23, 2025, according to Artificial Analysis.

What is MiniMax M2.1 good at?

MiniMax M2.1 is behind the leaders in long context, instruction following, factuality, agentic tasks, reasoning, and coding. Too few results yet to rate safety, math, multimodal tasks, or multilingual tasks.

How much does MiniMax M2.1 cost?

MiniMax M2.1 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 330 priced models we track.

How many benchmarks has MiniMax M2.1 been tested on?

We track 71 results for MiniMax M2.1 on 53 benchmarks from 11 sources, 7 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does MiniMax M2.1 support?

OpenRouter lists tool calling, json mode, and reasoning for MiniMax M2.1.

About this record

Where MiniMax M2.1's numbers come from, and every name it appears under.

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

Where the results come from

Verification: 71 scores · 7 independently verified · 15 aggregator-attributed · 14 vendor cross-reference · 35 vendor-reported. How these tiers are assigned

From 11 sources on 7 sites. Hugging Face supplies 27 of them; the 7 independently verified results come from 4 sites. Bars are coloured by trust tier.

  • huggingface.co27
  • api.llm-stats.com22
  • artificialanalysis.ai15
  • raw.githubusercontent.com4
  • datasets-server.huggingface.co1
  • labs.scale.com1
  • simple-bench.com1

Also known as

minimax-2.1minimax-m2p1minimax-m2-1minimax-m2.1-preview