MiniMax-01

Open Source
By MiniMaxReleased 2025-01-154M context456B (45.9B active) parameters
Open SourceAPI Available

Benchmark Scores

76.7/100
MMLU82.5
HumanEval78
Math66.5
Reasoning79
Coding77.5

About MiniMax-01

MiniMax-01 features a 4M token context window using a novel lightning attention mechanism. Its hybrid architecture enables efficient processing of extremely long sequences at lower cost than competing approaches.

Strengths

  • +4M token context window
  • +Lightning attention architecture
  • +Open source
  • +Very long document processing

Weaknesses

  • -Less known globally
  • -Complex architecture
  • -Fewer benchmark validations

Pricing

Per 1M tokens

Input$0.20
Output$1.10

Quick Facts

Context Window
4M
Parameters
456B (45.9B active)
Release Date
2025-01-15
Category
Open Source

Compare Models

See how MiniMax-01 stacks up against other models

Try MiniMax-01 on Vincony

Compare MiniMax-01 with 400+ other AI models side-by-side. Run AI debates, generate images and videos, and find the best model for your task.

Visit Vincony.com