MiMo V2.6 Pro
LabsWorth a try for long-horizon agentic coding.
- Artificial Analysis Intelligence Index
- 46
- #1 of 7 models we serve
- DeepSWE
- 71.9%* reported by the publisher
- #2 of 6 models we serve
- Terminal-Bench 2.1
- 89.9%* reported by the publisher
- #2 of 7 models we serve
- Context window
- 1M
- 128K max output
How it ranks
Against every model we serve, best first. Starred scores are reported by the publisher on its model card; the rest are independent measurements. About the scores
- MiMo V2.6 Pro (Labs)46
- GLM 5.345
- Kimi K344
- GLM 5.3 Flash42
- DeepSeek V4.1 Flash39
- DeepSeek V4 Flash34
- Umans Flash18
- DeepSeek V4.1 Flash74.2%* reported by the publisher
- MiMo V2.6 Pro (Labs)71.9%* reported by the publisher
- GLM 5.369%
- Kimi K368.5%
- GLM 5.3 Flash63.4%
- DeepSeek V4 Flash54.4%* reported by the publisher
Not published: Umans Flash.
- DeepSeek V4.1 Flash90.6%* reported by the publisher
- MiMo V2.6 Pro (Labs)89.9%* reported by the publisher
- Kimi K385%
- GLM 5.3 Flash84.3%
- GLM 5.383.9%
- DeepSeek V4 Flash78.7%
- Umans Flash44.9%
Pricing
Free with a Labs seat while the experiment runs. Seats are limited and rotate, and an experiment is not meant for production work.
Specs
- Context window
- 1M tokens
- Max output
- 128K tokens
- Vision
- Yes
- Tool calling
- Yes
- Reasoning
- Adjustable (none, low, medium, high)
- Default effort
- high
- Weights
- FP8 on Hugging Face
- Stage
- Labs experiment
Quickstart
Pick your tool; the command already carries this model's id. Calls need a Labs seat on your account while the experiment runs. More tools (Pi, Copilot, Zed, omp) in the setup guides.
OpenCode
Built inLaunch OpenCode on this model with the umans CLI. Umans AI is also built into OpenCode: run /connect and pick it.
curl -fsSL https://api.code.umans.ai/cli/install.sh | bash # once
umans opencode --model umans-mimo-v2.6-pro-lab
