mimo-v2.6-flash open weights and deployment

Status: Available. Xiaomi published the MiMo-V2.6-Flash-RL weights, which need custom code, under the MIT License declared in the model card, with SGLang and vLLM launch instructions. Xiaomi's pages do not say in so many words that this checkpoint is the one behind the mimo-v2.6-flash API.

Agent deployment prompt

You are my deployment agent. Deploy the official XiaomiMiMo/MiMo-V2.6-Flash-RL weights as a private, OpenAI-compatible service on this host. Use the publisher's model card as the source of truth: https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL.

Preflight
- Inspect the OS, GPU model and count, free VRAM, driver and CUDA versions, free disk space, and available ports. The model card's launch commands use tensor parallelism but state no minimum GPU; estimate whether the official weights and the requested context fit with safe headroom.
- The repository ships custom model code that needs trust-remote-code. Pin the exact revision you download, list the Python files it would execute, and ask me to approve them before enabling trust-remote-code.
- If the host cannot safely run the official weights, stop before installing or launching anything. Explain the limiting resource and propose a separately reviewed alternative; do not substitute a community quantization automatically.

Deployment
- If preflight passes and I have approved the remote code, create an isolated environment and follow the model card's SGLang or vLLM launch instructions, including its reasoning-parser and tool-call-parser settings. The model card warns that a stable vLLM release may lag behind; use the image or version it names rather than an untested one. Generate an API key and store it in a protected environment variable. Bind to 127.0.0.1 first; do not expose the endpoint publicly.
- Keep the publisher's sampling defaults (temperature 1.0, top_p 0.95). Start with a context length the preflight supports; the model supports up to 1,048,576 tokens.

Verification and handoff
- Send a real request to /v1/chat/completions and confirm that the model responds. Report the exact launch command, installed versions, model revision, local endpoint, test result, measured GPU memory and log location. Identify any reverse-proxy or firewall changes that still need my approval. Never print the API key or other secrets.
License
MIT (declared in the model card metadata)
Hardware note
Not published by Xiaomi beyond the tensor-parallel launch settings in the model card; the Agent recipe performs a hardware preflight.
Recommended runtime
SGLang or vLLM
Verified date
2026-09-24

Official repository

Published model card

Parameter count
309B total, 15B active per the model card; the technical report states 310B total
Context window
1M tokens (1,048,576)
Maximum output
128K tokens (131,072, including reasoning tokens)
Input/output modalities
Text, image, video and audio input; text output

Official sources