This is a quantized (MXFP4) version of the MiniMax-M3 model, optimized for efficient inference. It supports image and video inputs, tool calling via XML format, and configurable thinking modes. The model is distributed on Hugging Face and can be used with standard transformer libraries or specialized inference tools.
In the Foundation models & chat space, MiniMax M3 takes a focused approach. It focuses on running advanced multimodal reasoning models efficiently on consumer or edge hardware. It is built as an open-source project for developers and researchers. MiniMax M3 is open source under the Apache-2.0 license. It runs on the web, the command line, and API.
Behind MiniMax M3 is AMD, and the product first shipped in 2023. Development happens publicly on GitHub with 86.6k stars and 3k commits in the last 90 days. Key capabilities include Multimodal Input, Tool Calling, and Thinking Mode.
Latest indexed changes and source events
amd/MiniMax-M3-MXFP4 verified by the PulseGate indexer
Other apps tracked under the same category.
