
CORU-RRR PRO MEX
A micro AI model, extended through developer innovation
13 followers
A micro AI model, extended through developer innovation
13 followers
CORU is a research-stage micro language model exploring how developer-built architecture, APIs, tools and libraries can extend a tiny model's practical usefulness. Designed for experimentation by indie developers, startups and automation builders—not a claim of LLM-level capability.
This is the 2nd launch from CORU-RRR PRO MEX. View more

CORU-10M-ACT
Launching today
An open-source, 10M-parameter automation action-routing model that transforms natural-language intent into structured tool, action, and parameter proposals. Includes PyTorch inference for CPU and compatible TPU/XLA environments.




Free
Launch Team / Built With

Render Sandboxes: a computer for every agentIsolated compute, persistent state and a real filesystem
Promoted






Hello Product Hunt! 👋
We're excited to introduce CORU-10M-ACT, a lightweight automation action-routing model developed by RRR PRO MEX.
CORU is designed to bridge natural-language instructions and structured application automation. It identifies the intended action and prepares structured parameters that a host application can validate and process.
Unlike a general-purpose conversational model, CORU focuses specifically on automation action routing.
A few details:
• 9,995,264 parameters
• 55 supported actions across 23 namespaces
• CPU and compatible TPU/XLA inference support
• Open-source under the Apache License 2.0
CORU proposes structured actions; it does not independently execute external operations. Validation, authorization, user confirmation, and execution remain under the control of the host application.
Explore the model:
https://huggingface.co/rrrpromex/CORU-10M-ACT
Explore the source runtime:
https://github.com/rrrpro-ai/CORU-10M-ACT
We built CORU to explore a more focused and lightweight approach to application automation.
We would love to hear your feedback:
What automation workflows would you like to see supported by CORU?
Thank you for checking out our project!