MiniMax: MiniMax-01

Text input Image input Text output
Author's Description

MiniMax-01 is a combines MiniMax-Text-01 for text generation and MiniMax-VL-01 for image understanding. It has 456 billion parameters, with 45.9 billion parameters activated per inference, and can handle a context of up to 4 million tokens. The text model adopts a hybrid architecture that combines Lightning Attention, Softmax Attention, and Mixture-of-Experts (MoE). The image model adopts the “ViT-MLP-LLM” framework and is trained on top of the text model. To read more about the release, see: https://www.minimaxi.com/en/news/minimax-01-series-2

Key Specifications
Cost
$$$
Context
1M
Parameters
456B (Rumoured)
Released
Jan 14, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Top P Temperature Max Tokens
Performance Summary

MiniMax-01, created by minimax on January 14, 2025, is a sophisticated AI model combining text generation and image understanding capabilities, featuring a substantial 456 billion parameters. Its performance profile indicates competitive response times, ranking in the 50th percentile for speed across six benchmarks, suggesting it performs comparably to many models in terms of processing duration. In terms of cost, MiniMax-01 generally offers cost-effective solutions, placing in the 63rd percentile for price. A standout feature is its exceptional reliability, achieving the 100th percentile, indicating minimal technical failures and consistent provision of usable responses. Across benchmark categories, MiniMax-01 demonstrates strong performance in specific areas. It achieved perfect accuracy in the Ethics (Baseline) benchmark, highlighting its robust ethical reasoning capabilities and positioning it as the most accurate model at its price point and among models of similar speed. It also showed high accuracy in Email Classification (99%) and General Knowledge (99%), indicating strong classification and broad knowledge retention. While its Instruction Following (65%) and Reasoning (66%) scores are moderate, its Coding (81%) performance is solid. The model's hybrid architecture, combining Lightning Attention, Softmax Attention, and MoE for text, and a "ViT-MLP-LLM" framework for image understanding, likely contributes to its versatile capabilities.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.2
Completion $1.1

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Minimax
Minimax | minimax/minimax-01 1M $0.2 / 1M tokens $1.1 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by minimax