Process multiple modalities
The Pro and Flash models handle text, images, audio, and video.
Productivity / product dossier
MiMo-V2.6
Open omnimodal models with 1M context, plus training code, RL environments and a technical report.
Product brief
What MiMo-V2.6 does.
MiMo-V2.6 is Xiaomi’s open omnimodal model family for long-horizon agent work. Pro and Flash handle text, images, audio and video with 1M context, while Xiaomi is also releasing the technical report, RL environments and training code behind the public post-training run.
Why we selected it
A technically substantive open model release: omnimodal, 1M-context models accompanied by training code, RL environments, and a technical report. It is especially relevant amid current interest in systems for long-horizo
- Best for
- Open models
- Category
- Productivity
- Daily picks
- 1
- First selected
- 2026-09-22
Product preview saved with our daily selection
Capability scan
What it can help with.
Only capabilities supported by the product information we collected are listed here.
Support long-context work
The models are described as having a 1M context window for long-horizon agent work.
Provide public training materials
Xiaomi is releasing a technical report, RL environments, and training code behind its public post-training run.
Best-fit use cases
Build or evaluate agents that need to work across text, images, audio, and video.
Study the materials behind Xiaomi's public post-training run.
Explore long-horizon agent workflows using models with a stated 1M context window.
FAQ
Before you open it.
What is MiMo-V2.6?
MiMo-V2.6 is Xiaomi's open omnimodal model family for long-horizon agent work.
Which input types do the Pro and Flash models handle?
They handle text, images, audio, and video.
What context length is stated for MiMo-V2.6?
The description states that Pro and Flash have 1M context.
What is being released alongside the models?
Xiaomi is releasing a technical report, RL environments, and training code for the public post-training run.