MiniMax Releases M3 Open Weights with Modular Day-Zero Cloud Launch

MiniMax

Jun 13, 2026

MiniMax released M3 open weights, a multimodal model featuring a 1-million-token context window and a new sparse attention architecture. Modular is a day-zero launch partner, providing kernel-optimized inference on its cloud. The model, which uses approximately 428 billion total parameters and 23 billion activated parameters, is designed for frontier coding and long-horizon agentic tasks.

Arena; MiniMax M3 is in Agent Arena.; ARENA.AI/AGENT — MiniMax M3 model integration into the Agent Arena platform for performance evaluation and benchmarking purposes.

View the full update on minimax.io

MiniMax (official)

@MiniMax_AI1d ago

M3 open weight just dropped and it's live on @Modular cloud on day zero with up to a 1M-context and MSA architecture kernel-to-cloud optimization is exactly what M3 needs glad to have @Modular with us from the start

695

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

Keep reading

MiniMax brings M3 to local PCs with 1M context open weights

MiniMax announced that its M3 model is joining the NVIDIA and Microsoft local LLM lineup, with weights releasing to the community within 10 days. The move brings high-capacity multimodal reasoning and coding capabilities directly to local hardware.

Together AI powers MiniMax M3 with 1M context and sparse attention

Together AIJun 3

Together AI powers MiniMax M3 with 1M context and sparse attention

Together AI is now powering inference for MiniMax M3, a multimodal model featuring a 1-million-token context window. The model uses a new sparse attention architecture to process massive datasets with significantly lower computational overhead than previous-generation models.

OpenRouter adds MiniMax-M3 with 1M context for multimodal agentic coding

OpenRouterJun 1

OpenRouter adds MiniMax-M3 with 1M context for multimodal agentic coding

OpenRouter integrated MiniMax-M3, an open-weight multimodal model featuring a 1-million-token context window and specialized sparse attention. By reducing long-context compute costs by 95%, the model enables persistent agentic workflows across massive codebases and video files.

Fireworks AI hosts MiniMax M3 with 15x faster long context decoding

Fireworks AIJun 4

Fireworks AI hosts MiniMax M3 with 15x faster long context decoding

Fireworks AI is now powering inference for MiniMax M3, a multimodal model featuring a novel sparse attention architecture. The partnership enables 15.6x faster decoding at 1-million-token context, making real-time agentic workflows viable at scale.