Back to all models

MiniMax M3

MiniMax M3 is a multimodal foundation model with 1M token context, native multimodal understanding, and MiniMax Sparse Attention (MSA) for efficient long-context inference.

minimax-m3
STABLEGet StartedView uptime
1,048,576 context
Starting at $0.30/M (50% off) input tokens
Starting at $1.20/M (50% off) output tokens
Streaming
Vision
Tools
Reasoning
JSON Output
50% offthis model via minimax—6d 9h 39m remaining

Select Provider

All Providers for MiniMax M3

LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.

MiniMax
Context: 1.0M50% off
Input
$0.6$0.3
50% off
/M tokens
Cached
$0.12$0.06
50% off
/M tokens
Output
$2.4$1.2
50% off
/M tokens
Get Started