MiniMax
Pro

MiniMax M3

minimax-m3

MiniMax's flagship model — code, reasoning and analysis with a 1 million token context.

CodeReasoningAnalysis

Technical Specifications

API Identifierminimax-m3
PlansPro
Maximum Context1 000 000 tokens
Vision (images)No
Native Web AccessNo
Image GenerationNo
Input Cost$1.50 / M tokens
Output Cost$6.00 / M tokens

Overview

MiniMax M3 is MiniMax's most advanced model, succeeding M2.7. It combines excellent code, reasoning and analysis capabilities with the signature context window of the MiniMax family: 1 million tokens. Mingo reserves it for the Pro plan for projects requiring both high quality and the ability to process huge amounts of information in a single request.


Strengths

MiniMax flagship1M token contextCode and reasoningIn-depth analysis

Ideal Use Cases

Whole codebase analysis

Analyze and understand entire software projects in a single request.

Massive documentation synthesis

Summarize and cross-reference very large sets of documents.

Reasoning over large volumes

Solve complex problems requiring a lot of context.

Large-scale data analysis

Process and analyze very extensive datasets.

Pricing

Input Tokens
$1.50
per million tokens
Output Tokens
$6.00
per million tokens

Related Models

MiniMax M2.7 High Speed
Basic — Version rapide
MiniMax M2.7
Basic — Génération précédente
Qwen3.7 Max
Pro — Alternative analyse

Frequently Asked Questions

Does MiniMax M3 replace MiniMax M2.7?
MiniMax M3 is a new, more capable generation reserved for the Pro plan, while M2.7 and its High Speed variant remain available on Basic.
Does MiniMax M3 really have 1M tokens of context?
Yes, like the entire MiniMax family, it has a 1 million token context window.
When does Mingo choose MiniMax M3?
For the most demanding code, reasoning or analysis tasks requiring both high quality and a very large context.

Use MiniMax M3 with Mingo

Mingo routes your most demanding projects requiring a large context to MiniMax M3 on the Pro plan.

Try Mingo for free