MiniMax M3
MiniMax M3 is a June 2026 open-weight multimodal model for coding, agentic workflows, computer use, and long-context work.
Provider
MiniMax
Model family
MiniMax M
Open-weight multimodal LLM
Cost tier
Flagship
Status
Current
Release Jun 1, 2026
Why teams choose it
Evaluate M3 on end-to-end computer-use and repository tasks because multimodal agent
quality depends on action reliability, not only text benchmarks.
Keep permissions narrow for computer-use workflows and require explicit approval for side effects.
Tradeoffs to know
- M3 is not the only model called M3 in technology markets; use the MiniMax name in procurement and telemetry.
- Verify checkpoint license, host requirements, regional access, and tool safety before rollout.
Benchmarks
No independent or vendor benchmark scores in this catalog entry.
Technical specs
- Inputs
- text, image, video
- Outputs
- text
- License
- Open weights; verify the MiniMax license for the selected checkpoint
- Capabilities
- agentic coding
- computer use
- image understanding
- video understanding
- long context
- open-weight deployment
MiniMax M3 FAQ
What is MiniMax M3?
MiniMax M3 is a June 2026 open-weight multimodal model for coding, agentic workflows, computer use, and long-context work. MiniMax documents a one-million-token context window, native image and video understanding, and deployment through hosted or downloadable model paths.
When does MiniMax M3 fit best?
Multimodal coding agents
What should teams watch out for with MiniMax M3?
M3 is not the only model called M3 in technology markets; use the MiniMax name in procurement and telemetry.
Explore next
Models, tools, and comparisons that connect to this reference.