Get Started with Cleo

Welcome to the Cleo ecosystem developed by Inserloft.

Cleo is focused on:

  • Programming
  • Edge AI
  • Mobile inference
  • Efficient deployment
  • Modern AI systems

The ecosystem is designed around specialized AI models optimized for different real-world environments and workloads.


What is Cleo?

Cleo is the AI ecosystem created by Inserloft.

Rather than building a single universal model for every possible task, Cleo is designed as a family of specialized AI systems optimized for:

  • efficiency,
  • reasoning,
  • deployment,
  • and scalability.

Our goal is to create AI systems that are practical, deployable, and accessible across modern hardware environments.


The Cleo Model Family

Cleo NaNo is a lightweight AI model focused on:

  • programming,
  • edge AI,
  • mobile inference,
  • and efficient local deployment.
  • Lightweight architecture
  • Optimized inference
  • Programming-oriented reasoning
  • Fast deployment
  • Efficient local execution
  • 22M → 52M parameters
  • Better code understanding
  • Improved context retention
  • Conversational stability
  • Faster inference optimization

https://huggingface.co/Inserloft/NaNo


Cleo Kyro is an advanced reasoning model designed for:

  • broader intelligence,
  • deeper context understanding,
  • and stronger conversational reasoning.
  • Advanced reasoning
  • Conversational intelligence
  • Context understanding
  • Scalable architecture
  • Modern optimization

https://huggingface.co/Inserloft/Kyro


Real-World Deployment

The Cleo ecosystem is designed for modern AI deployment across multiple environments.

AI systems optimized for technical workflows and software development.

Lightweight deployment for local and edge environments.

Optimized models designed for modern mobile hardware and applications.

Developer-focused experimentation and open-source releases.


Open Source

Inserloft currently supports open-source releases for:

  • experimentation,
  • local inference,
  • deployment workflows,
  • and developer tooling.

Available formats may include:

  • GGUF
  • safetensors
  • PyTorch checkpoints

Our goal is to help developers:

  • build,
  • experiment,
  • deploy,
  • and integrate AI systems more efficiently.

Ecosystem Overview

ModelFocus
Cleo NaNoLightweight programming and edge AI
Cleo KyroAdvanced reasoning and intelligence

This approach allows Inserloft to optimize AI systems for different workloads instead of relying on a single oversized model.


What’s Next

Development of the Cleo ecosystem is ongoing.

Future updates will continue improving:

  • reasoning capabilities,
  • context handling,
  • inference speed,
  • deployment tooling,
  • mobile optimization,
  • and programming intelligence.

Additional documentation, APIs, and developer tools will continue expanding over time.


Useful Links

https://huggingface.co/Inserloft

https://inserloft.com/Blog

https://inserloft.com/Pixel/Community