llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsMiniMax M2

MiniMax model product

MiniMax M2

MiniMax M2 is an open-source large language model by MiniMax, built for agents and coding tasks. It delivers state-of-the-art tool use, reasoning, and search performance while maintaining exceptional cost-efficiency and speed, priced at just 8% of Claude 3.5 Sonnet’s cost and running at nearly double its inference speed (≈100 TPS). Designed for end-to-end agentic workflows, it excels at long-chain tool calling across Shell, Browser, Python, and other MCP tools. While slightly behind top overseas models in programming, it ranks among the best domestic models and top five globally on the Artificial Analysis benchmark. M2 powers the MiniMax Agent platform, available in Lightning Mode for fast tasks and Pro Mode for complex multi-step reasoning, and its weights, API, and deployment guides are freely available on Hugging Face, vLLM, and SGLang.

Updated Aug 10, 2026. Default version: MiniMax M2

Compare
LLMBoard score44.9MiniMax M2
Coverage80%7 benchmark families
Context window196.6KTokens
Official input price$0.30MiniMax (minimax.io) API

On this page

  • Specification
  • Capability
  • Benchmarks
  • Arena
  • Pricing
  • Versions
  • About
  • Compare
  • Similar models
  • FAQ

Model specification

Structured fields from the published default version.

Version
MiniMax M2
Released
Oct 27, 2025
Knowledge cutoff
Unknown
Parameters
230B
Context window
196.6K
Max output
128K
Inputs
text
Outputs
text
Open weights
Yes
License
MIT

Capability profile

This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.

MiniMax M2 category scores

Benchmark results

Published benchmark records for the scored version MiniMax M2.

16 rows
Columns

Show columns

IF0.7220.0%CAug 7, 2026
AA-Index0.6330.0%CAug 7, 2026
Terminal-Bench0.542587.5%CAug 7, 2026
LiveCodeBench0.857394.4%CAug 7, 2026
Multi-SWE-Bench0.45620.0%CAug 7, 2026
Tau-bench0.85620.0%CAug 7, 2026
BrowseComp-zh0.5111316.7%CAug 7, 2026
SciCode0.417185.9%CAug 7, 2026
Tau2 Telecom0.9173552.9%CAug 7, 2026
SWE-bench Multilingual0.6273421.2%CAug 7, 2026
MMLU-Pro0.83912970.3%CAug 7, 2026
BrowseComp0.4505814.0%CAug 7, 2026
SWE-Bench Verified0.76210440.8%CAug 7, 2026
AIME 20250.87511434.5%CAug 7, 2026
Humanity's Last Exam0.1769217.6%CAug 7, 2026
GPQA0.89423359.9%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

66 rows
Columns

Show columns

webdevwebdev-html891306.86,473N/AAug 6, 2026
webdevoverall901296.76,502N/AAug 6, 2026
webdevwebdev901296.76,502N/AAug 6, 2026
text factualityindustry business and management and financial operations1251375.6758N/AAug 6, 2026
text factualityindustry entertainment and sports and media1251322.6602N/AAug 6, 2026
text factualitycoding1301418.8873N/AAug 6, 2026
text factualityindustry writing and literature and language1311317.4846N/AAug 6, 2026
text factualityhard prompts english1331402.91,064N/AAug 6, 2026
text factualityinstruction following1331346.21,126N/AAug 6, 2026
text factualitylonger query1341359.1900N/AAug 6, 2026
text factualityindustry software and it services1351406.81,433N/AAug 6, 2026
text factualityhard prompts1411389.23,399N/AAug 6, 2026
text factualityenglish1441387.42,938N/AAug 6, 2026
text factualityexclude ties1461334.74,819N/AAug 6, 2026
text factualitynon english1471335.23,137N/AAug 6, 2026
text factualityoverall1491362.86,692N/AAug 6, 2026
textmulti turn1571367.21,116N/AAug 6, 2026
textchinese1621384.3278N/AAug 6, 2026
textindustry mathematical1641367.2286N/AAug 6, 2026
textindustry medicine and healthcare1651365.8319N/AAug 6, 2026
texthard prompts english1691375.21,902N/AAug 6, 2026
textpolish1691316.1266N/AAug 6, 2026
textcoding1701373.71,545N/AAug 6, 2026
textindustry software and it services1701376.52,468N/AAug 6, 2026
text style controlindustry mathematical1701369.6286N/AAug 6, 2026
textspanish1731319.0173N/AAug 6, 2026
text style controlchinese1731386.9278N/AAug 6, 2026
text style controlmulti turn1731363.91,116N/AAug 6, 2026
texthard prompts1741359.13,691N/AAug 6, 2026
textinstruction following1741333.71,983N/AAug 6, 2026
textindustry business and management and financial operations1771338.01,314N/AAug 6, 2026
textrussian1771328.3323N/AAug 6, 2026
textindustry legal and government1811346.1429N/AAug 6, 2026
textlonger query1811335.61,582N/AAug 6, 2026
textenglish1821364.23,363N/AAug 6, 2026
textoverall1831342.46,859N/AAug 6, 2026
textexclude ties1831299.34,998N/AAug 6, 2026
text style controlhard prompts english1831387.71,902N/AAug 6, 2026
text style controlmath1851353.7320N/AAug 6, 2026
text style controlpolish1851310.1266N/AAug 6, 2026
textexpert1861320.9256N/AAug 6, 2026
textindustry life and physical and social science1871349.31,108N/AAug 6, 2026
textmath1871351.0320N/AAug 6, 2026
text style controlhard prompts1871368.73,691N/AAug 6, 2026
textindustry entertainment and sports and media1881302.71,175N/AAug 6, 2026
text style controlenglish1891370.83,363N/AAug 6, 2026
text style controlindustry software and it services1891386.02,468N/AAug 6, 2026
text style controlindustry business and management and financial operations1901349.51,314N/AAug 6, 2026
textnon english1911315.83,448N/AAug 6, 2026
text style controlindustry medicine and healthcare1911364.0319N/AAug 6, 2026
text style controlinstruction following1911338.91,983N/AAug 6, 2026
text style controlrussian1921332.9323N/AAug 6, 2026
text style controlspanish1931310.6173N/AAug 6, 2026
text style controlindustry legal and government1951357.0429N/AAug 6, 2026
text style controlexclude ties1991303.64,998N/AAug 6, 2026
text style controlexpert1991337.9256N/AAug 6, 2026
text style controloverall2001346.06,859N/AAug 6, 2026
text style controllonger query2011343.21,582N/AAug 6, 2026
textcreative writing2021290.6892N/AAug 6, 2026
text style controlcoding2021384.71,545N/AAug 6, 2026
textindustry writing and literature and language2031299.11,467N/AAug 6, 2026
text style controlindustry entertainment and sports and media2031302.51,175N/AAug 6, 2026
text style controlindustry life and physical and social science2031356.61,108N/AAug 6, 2026
text style controlnon english2061316.33,448N/AAug 6, 2026
text style controlcreative writing2261288.5892N/AAug 6, 2026
text style controlindustry writing and literature and language2301302.11,467N/AAug 6, 2026

Pricing

Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.

Official API
$0.30 input, $1.2 output per 1M
Official provider
MiniMax (minimax.io)
Lowest third-party
From $0.17 input, $1.53 output per 1M via NanoGPT
Tracked offerings
8
8 rows
Columns

Show columns

MiniMax Token Plan (minimax.io)MiniMax-M2globalN/AN/A196.6KAug 7, 2026
MiniMax Token Plan (minimaxi.com)MiniMax-M2globalN/AN/A196.6KAug 7, 2026
NanoGPTMiniMax-M2global$0.17$1.53200KAug 7, 2026
LLM Gatewayminimax-m2global$0.20$1196.6KAug 7, 2026
MiniMax (minimax.io)MiniMax-M2global$0.30$1.2196.6KAug 7, 2026
MiniMax (minimaxi.com)MiniMax-M2global$0.30$1.2196.6KAug 7, 2026
302.AIMiniMax-M2global$0.33$1.321MAug 7, 2026
Cortecsminimax-m2global$0.349$1.41400KAug 7, 2026

Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.

MiniMax M2 versions

All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.

1 rows
Columns

Show columns

MiniMax M2Oct 27, 202544.9230B196.6K128KYesMIT

What is MiniMax M2?

A concise description based on the published model registry.

MiniMax M2 is an open-source large language model by MiniMax, built for agents and coding tasks. It delivers state-of-the-art tool use, reasoning, and search performance while maintaining exceptional cost-efficiency and speed, priced at just 8% of Claude 3.5 Sonnet’s cost and running at nearly double its inference speed (≈100 TPS). Designed for end-to-end agentic workflows, it excels at long-chain tool calling across Shell, Browser, Python, and other MCP tools. While slightly behind top overseas models in programming, it ranks among the best domestic models and top five globally on the Artificial Analysis benchmark. M2 powers the MiniMax Agent platform, available in Lightning Mode for fast tasks and Pro Mode for complex multi-step reasoning, and its weights, API, and deployment guides are freely available on Hugging Face, vLLM, and SGLang.

Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.

Data snapshot: 2026-08-07. Editorial model content is not available in the backend.

MiniMax M2 vs nearby models

Open a comparison with the three ranked models immediately above and below this model.

MiniMax M2vsGemma 3 27BMiniMax M2vsNemotron 3 NanoMiniMax M2vsGranite 3.3 8BMiniMax M2vsLlama 3.1 Nemotron Nano 8BMiniMax M2vsDeepSeek-R1MiniMax M2vsQwen2.5 14B

Models similar to MiniMax M2

Recommendations prioritize the same model type and family, then the closest published LLMBoard score.

#89+3.6
MI

MiniMax M1 40K

MiniMax

48.5 LLMBoard

DetailsCompare
#68+10.0
MI

MiniMax M2.1

MiniMax

54.8 LLMBoard

DetailsCompare
#62+11.5
MI

MiniMax M1 80K

MiniMax

56.3 LLMBoard

DetailsCompare
#98+0.2
IB

Granite 3.3 8B

IBM

45.0 LLMBoard

DetailsCompare
#97+0.2
NV

Nemotron 3 Nano

NVIDIA

45.1 LLMBoard

DetailsCompare
#96+0.3
GO

Gemma 3 27B

Google

45.1 LLMBoard

DetailsCompare

FAQ

Common questions about MiniMax M2.

When was MiniMax M2 released?

MiniMax M2's default version was released on Oct 27, 2025.

How much does MiniMax M2 cost?

MiniMax M2's official API price is $0.30 per million input tokens and $1.2 per million output tokens via MiniMax (minimax.io). The lowest tracked third-party offer starts at $0.17 input and $1.53 output via NanoGPT.

Who created MiniMax M2?

MiniMax M2 is published under MiniMax in the model registry.

What is the context window for MiniMax M2?

The default version has a 196.6K token context window.

Is MiniMax M2 open weight?

Yes. The default version is marked as open weight under MIT.

How many API providers offer MiniMax M2?

8 published provider offerings are linked to the default version.

What models should I compare MiniMax M2 with?

Nearby ranked alternatives include Gemma 3 27B, Nemotron 3 Nano, Granite 3.3 8B.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai