Nomi AI by Lazy Loop Studio

Small, fast & indie‑built language models for real people, not just benchmarks.

🌱 About Nomi

Nomi is a series of compact, high‑quality language models designed to run on everyday hardware while still feeling surprisingly capable. Instead of relying on massive data centers, Nomi focuses on efficient 3B models that you can run locally, privately, and with minimal setup.

Our models are primarily optimized for clean structure, solid Markdown, and useful code, making them a great fit for developers, indie projects, and anyone who wants full control over their AI stack.

🧩 Indie Spirit

Nomi AI is a small, independent project driven by a passion for small but strong models, not by a corporate budget. We experiment with modern base models, fine‑tune them on carefully selected datasets, and try to squeeze as much “brain” as possible out of just 3B parameters.

If you care about local, open, and hackable AI tools, you are in the right place.

🔧 What We Build

• 🧠 3B models that feel like larger models (e.g., Nomi‑1.0, Nomi‑1.1).
• 📝 Assistants focused on structured reports, Markdown, and developer workflows.
• 🌍 Models with strong German/English support for European developer setups.

Whenever possible, our models are released under the Apache‑2.0 license, so you can freely use, modify, and integrate them into your own projects.

Our Models

• Nomi‑1.0‑3B – Base model with core instruction tuning and strong Markdown + German/English support
  👉 https://huggingface.co/LL-Nomi-AI/Nomi-1.0-3b

• Nomi‑1.1‑3B (GGUF) – Refined behavior, better formatting, and improved multilingual abilities, optimized for local inference
  👉 https://huggingface.co/LL-Nomi-AI/Nomi-1.1-GGUF

🗒️ Our Plan

Right now, we are fully focused on Nomi. Our goal is to develop the series up to Version 2.0 and make it competitive with much larger models:

• Nomi 1.0 → Basic instruction tuning
• Nomi 1.1 → Improved behavior plus some code and math
• Next step → Stronger coding capabilities
• After that → Stronger math and reasoning
• One step further → Agentic, tool‑using tasks that can stand alongside bigger models

We are also thinking about making vision and reasoning models

🤝 Join the Journey

We welcome feedback, benchmarks, pull requests, and experiments using Nomi in your own projects.
If you are building on laptops, mini PCs, or homelabs and want something fast and lightweight, give Nomi a try and tell us what you built with it.

Check out our homepage: https://lazyloopstudio.netlify.app