RT @jun_song: My goal on Minimax M3 post-train :
MiniMax AI has released MiniMax M3, a new open-source model with 428 billion parameters and 1 million tokens of context. This model aims for uncensored output and performance comparable to Gemini 3.1 Pro. It can be run locally, with dynamic 2-bit GGUF requiring 138GB of RAM/VRAM and 3-bit requiring 165GB. AI
IMPACT This open-source release with a large context window and competitive performance could spur further innovation in local LLM deployment.