DataLearner logo
MI

MiniMax M3

Multimodal modelCoding modelMiniMax M3

MiniMax M3

Release date: 2026-06-01Updated: 2026-08-22Views: 6,428
Live demoGitHubHugging FaceCompare
Parameters
428B
Context length
1M
Multilingual
Supported
Reasoning ability
4/5

MiniMax M3 is a multimodal model from MiniMaxAI, released on 2026-06-01. It accepts text, image, and video input and returns text output. The cataloged parameter count is 428B, with 23B active parameters per inference. The recorded context window is 1M, and the recorded maximum output is 512K. Cataloged capabilities include Reasoning model, Multilingual, and Coding model. The checkpoint is listed under the Minimax Modified Mit license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

MiniMax M3

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Mode (Default)Standard Mode
Context length
1M tokens
Max output length
512K tokens
Model type
Multimodal model
Modality (in / out)
Text, Image, Video → Text
Release date
2026-06-01
Model file size
No data
MoE architecture
Yes
Total params / Active params
428B / 23B
Knowledge cutoff
No data
MiniMax M3

Open source & experience

Code license
MIT License
Weights license
MiniMax Modified MIT- Non-commercial only
GitHub repo
github.com/MiniMax-AI/MiniMax-M3
Hugging Face
huggingface.co/MiniMaxAI/MiniMax-M3
Live demo
agent.minimax.io
MiniMax M3

Official resources

Paper
Paper
DataLearnerAI blog
N/A
MiniMax M3

API details

API speed
3/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Learn about pricing modes
Standard
TypeConditionInputOutput
Text-¥2.10/ 1M¥8.40/ 1M
TextContext <= 524288¥4.20/ 1M¥16.80/ 1M
Turbo
TypeConditionInputOutput
Text-¥3.15/ 1M¥12.60/ 1M
TextContext <= 524288¥6.30/ 1M¥25.20/ 1M
Cache PricingPrompt Cache
TypeTTLWriteRead
Text-—¥0.420/ 1M

“—” means the modality is not billed in that direction, or the vendor has not published a price for it.

MiniMax M3

Benchmark Results

MiniMax M3 currently shows benchmark results led by AA-LCR (1 / 27, score 80.33), BrowseComp (12 / 54, score 83.50), SWE-Bench Pro - Public (15 / 59, score 59). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage
Internet

General Evaluation

1 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
81.31
132 / 270

Coding and Software Engineer

4 evaluations
Benchmark / mode
Score
Rank/total
Text Arena (Coding)
Standard Mode
1527.75
14 / 35
SWE-Bench Pro - Public
Thinking ModeTools
59
15 / 59
SciCode
Thinking Mode
45.37
10 / 14
PostTrain Bench
Thinking ModeTools
37
2 / 5

AI Agent - Information Search

1 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
Thinking ModeToolsInternet
83.50
12 / 54

General Knowledge

2 evaluations
Benchmark / mode
Score
Rank/total
LiveBench
Deep
70.02
40 / 115
AA Intelligence Index
Thinking Mode
45.40
8 / 12

Text Embedding

1 evaluations
Benchmark / mode
Score
Rank/total
Context Arena
Thinking Mode
51.15
87 / 126

Long Context

1 evaluations
Benchmark / mode
Score
Rank/total
AA-LCR
Thinking Mode
80.33
1 / 27

AI Agent - Tool Usage

3 evaluations
Benchmark / mode
Score
Rank/total
MCP-Atlas
Thinking ModeTools
74.20
24 / 40
OSWorld-Verified
Thinking ModeTools
70
19 / 26
Terminal-Bench 2.1
Thinking ModeTools
66
38 / 47

Productivity Knowledge

3 evaluations
Benchmark / mode
Score
Rank/total
GDPval-AA v2
Thinking ModeTools
1379.75
16 / 22
AA-Briefcase
Thinking ModeTools
1107.16
14 / 19
AA-AnalystAgent
Thinking ModeToolsInternet
10
11 / 12

Agent Level Benchmark

1 evaluations
Benchmark / mode
Score
Rank/total
τ³-Banking
Thinking ModeTools
15.26
10 / 12
View benchmark analysisCompare with other models

Compare with other models

MiniMax M3

Publisher

MiniMax M3

Model Overview

MiniMax M3 is a multimodal model from MiniMaxAI, released on 2026-06-01.

It accepts text, image, and video input and produces text output. Its cataloged capabilities include Reasoning model, Multilingual, and Coding model. The cataloged parameter count is 428B, with 23B active parameters per inference. The recorded context window is 1M, and the recorded maximum output is 512K.

The checkpoint is listed under the Minimax Modified Mit license, with the license link included in the model references. The page records 12 API pricing rules from MiniMaxAI; current provider pricing and conditions should be checked before deployment. The evaluation section contains 5 cataloged benchmark results with their recorded modes and scores. The page links 4 release, model-card, repository, or provider references for checking the underlying claims. Specifications, availability, and prices can change; undisclosed values are intentionally left unstated.

MiniMax M3

FAQ

What is MiniMax M3?

MiniMax M3 is a multimodal model from MiniMaxAI, released on 2026-06-01. It accepts text, image, and video input and returns text output. The cataloged parameter count is 428B, with 23B active parameters per inference. The recorded context window is 1M, and the recorded maximum output is 512K. Cataloged capabilities include Reasoning model, Multilingual, and Coding model. The checkpoint is listed under the Minimax Modified Mit license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

What input and output modalities does MiniMax M3 support?

The current model record lists text, image, and video as input and text as output.

What are the main recorded specifications for MiniMax M3?

The cataloged parameter count is 428B, with 23B active parameters per inference. The recorded context window is 1M, and the recorded maximum output is 512K. Fields without a source-backed value remain undisclosed.

Does MiniMax M3 have API pricing?

The page records 12 API pricing rules from MiniMaxAI; current provider pricing and conditions should be checked before deployment.

Are benchmark results available for MiniMax M3?

The evaluation section contains 5 cataloged benchmark results with their recorded modes and scores. Compare only results that use the same benchmark version and evaluation mode.

Is MiniMax M3 open source?

The checkpoint is listed under the Minimax Modified Mit license, with the license link included in the model references. Review the linked license text before commercial or derivative use.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code