MiniMax-M2.7 with self-improvement & research in OpenClaw

Snippet

Body

 




 

MiniMax-M2.7 is now available to run via Ollama's cloud. This new model is designed to excel at completing end-to-end tasks in
software engineering and office productivity while improving its character and emotional intelligence.
 
M2.7 is also good at self-improvement using data from the web. This allows assistants like OpenClaw to do research and learn new
skills.
 
To get started, download Ollama:
 

Download Ollama [https://ollama.com/download]

 
Next, launch OpenClaw with MiniMax-M2.7:
 
ollama launch openclaw --model minimax-m2.7:cloud
 
If you already have OpenClaw running with Ollama, switch to MiniMax-M2.7 cloud by using the /model command:
 
/model ollama/minimax-m2.7:cloud


 


Self-improvement & research

OpenClaw can now perform deep research tasks and use up-to-date web information to teach itself skills and best practices.
 
Example: financial reporting
 
Teach yourself how to write financial research reports by studying examples from top analysts. Write a PPT presentation and Word
document report on TSMC based on their latest earnings. Save your research framework as a skill so future reports follow the same
process.
 


Launch agents in non-interactive mode

Ollama can now launch non-interactive agents using the new --yes flag.
 
This enables running OpenClaw and other agents in the background without interruption on servers, scripts, and other environments
that do not support the OpenClaw gateway.
 
To launch a non-interactive OpenClaw agent, run:
 
ollama launch openclaw \
--model minimax-m2.7:cloud \
--yes -- agent \
--agent main \
--local \
--message "Prepare a pre-read for my next meeting"
 
 
If you have any feedback, reply directly to this email or join Ollama's Discord channel [https://discord.gg/ollama].
 
❤️ Ollama
 
You are receiving this email because you opted-in to receive updates from Ollama
Ollama, 744 High Street, Palo Alto, CA 94301
Unsubscribe [https://app.loops.so/unsubscribe/cmmxambr3ab5f01ay8xed4m89/608e8d28b49efc2859ad5adbd326772defeed9603120412d8fd9ac4bf7fbc4df4d0a853e412eacb41f03cd2dbede664aa8844d1b536278d4bcb93262]