---
title: README
emoji: ๐
colorFrom: pink
colorTo: purple
sdk: static
pinned: false
thumbnail: >-
https://cdn-uploads.huggingface.co/production/uploads/68a5d0966d33a07f8aad2e51/vEv2iwEIhlXC3sG_2Qn9m.png
short_description: Open-Source Small Language Model Research
---
Welcome to SupraLabs!
SupraLabs is a small independent research group focused on educational and experimental AI systems built with limited consumer hardware.
Our mission is to advance efficient and accessible AI by developing open-source Language and Multimodal Models, datasets, and tools that can run on a wide range of hardware.
---
## Latest releases
| Model | Description |
|---------|-------------|
| [Supra-1.5 Instruct (EXP)](https://huggingface.co/SupraLabs/Supra-1.5-50M-instruct-exp-gguf) | Instruction-tuned experimental release. [Demo](https://huggingface.co/spaces/SupraLabs/Supra1.5-50M-Instruct-Demo) |
| [Supra-1.0 Reasoning (50M)](https://huggingface.co/SupraLabs/Supra-50M-Reasoning) | First-generation reasoning model. [Demo](https://huggingface.co/spaces/SupraLabs/Supra-50M-Reasoning-Demo) |
| [Supra-1.5 Base (EXP)](https://huggingface.co/SupraLabs/Supra-1.5-50M-Base-exp) | Base model (v1.5 ยท Latest). |
| [Supra-50M-Instruct(EXP)](https://huggingface.co/SupraLabs/Supra-50M-Instruct) | First-generation chat model. |
| [Supra Title 50M (PRE)](https://huggingface.co/SupraLabs/supra-title-50M-pre-gguf) | First-generation chat title model built on SupraLabs' own foundation model. |
| [Supra-A2A-Nano](https://huggingface.co/SupraLabs/Supra-A2A-Nano-Exp) | Multimodal experimental Any-to-Any model |
| [All Models from SupraLabs](https://huggingface.co/SupraLabs/models) | View all the AI models from SupraLabs |
_Datasets_
| Dataset | Description |
|----------|-------------|
| [150K Chat Titles (Unfiltered)](https://huggingface.co/datasets/SupraLabs/chat-titles-unfiltered-150K) | 150K samples of unfiltered chat title data. |
| [115K Chat Titles (Filtered)](https://huggingface.co/datasets/SupraLabs/chat-titles-filtered-115K) | 115K filtered chat titles generated from the first user message. |
| [500ร Supra-1.0 Reasoning](https://huggingface.co/datasets/SupraLabs/SupraThink-Dataset-500x) | 500 reasoning traces from Qwen-3-1.7B used to train the first-generation reasoning model. |
_Blogs & Research_
| Resource | Description |
|-----------|-------------|
| [Blog](https://huggingface.co/spaces/SupraLabs/Blog) | Read our latest blog posts. |
| [Research](https://huggingface.co/spaces/SupraLabs/Research-SupraLabs) | Explore our research publications and findings. |
---
## Our Naming Conventions:
At SupraLabs, we use a strict naming system for our models to avoid confusion!
- **EXP (Experimental)** โ Early-stage release intended for testing and development. These models may be unstable, incomplete, or subject to significant changes.
- **Preview** โ A near-final release that provides an early look at the intended final model. Minor tuning, optimizations, or improvements may still be pending.
- **No Tag** โ Stable final release. Recommended for general use and represents the intended production version.
---
### Who are we?
_We are_
- @AxionLab-official
- @LH-Tech-AI
- @QyrouNnet-AI
- @User01110
- @MariusNocturnum
- @lesageethan
We are creating small open-source models for everyone.
Our models can be found in the [open SLM leaderboard](https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard) โ feel free to check out!
You can support us with a follow! ๐