---
title: Tiny-LM Leaderboard
emoji: 📊
colorFrom: blue
colorTo: green
sdk: static
pinned: true
sdk_version: 2.0.0
short_description: Every tiny LM, same eval harness, transparent benchmarks
---

A community leaderboard for sub-100M parameter language models. Includes CompactAI-O models, SupraLabs, and any other tiny LM with verifiable benchmarks.

Data is pull-request driven. Open an issue or PR to add your model.

## Methodology

All benchmarks use the CompactAI eval suite (sliding-window log-prob scoring). Methodology and reproduction scripts are linked from each model card.

## Links

- [CompactAI-O on HF](https://huggingface.co/CompactAI-O)
- [lh-tech.de comparison](https://lh-tech.de/ai/compare-tiny-models.html) — independent community page
- [Submit a model](https://huggingface.co/spaces/CompactAI-O/Tiny-LM-Leaderboard/discussions)