README.md

metadata

title: Model Rank
emoji: ⚡
colorFrom: indigo
colorTo: blue
sdk: gradio
sdk_version: 5.49.1
app_file: app.py
pinned: true
license: apache-2.0
short_description: Browse the most popular models by category (base, quant, FT)
tags:
  - huggingface-space
  - gradio-app
  - llm-discovery
  - model-ranking
  - fine-tune
  - quantized-models
  - open-source-ai
  - huggingface-models
  - discovery
  - recommendation

VANTA Research

Independent AI safety research lab specializing in cognitive fit, alignment, and human-AI collaboration

---

Overview

Model Rank enables users to apply several different filters in order to quickly and easily find models on Hugging Face

Features

• Categorized View: Browse models separated into Base Models, Fine-Tunes, and Quants
• Smart Filters: Narrow results by model family, license, and access (Open vs. Gated)
• Timeframe Control: Focus on freshly released models with Last Day, Week, Month, or 3 Month filters
• Hidden Gems Discovery: Surface under-the-radar models with strong like/download ratios and reproducibility signals
• Flexible Sorting: Toggle between downloads or likes to match your exploration style
• Clean Interface: Professional, easy-to-use Gradio experience with quick refresh controls
• Auto Refreshed Data: Pulls the latest metadata from Hugging Face Hub with lightweight caching to stay snappy

Categories

• Base Models: Foundation models and pre-trained base models
• Fine-Tunes: Models fine-tuned for specific tasks
• Quants: Quantized versions of models (GGUF, GPTQ, AWQ, etc.)

Usage

1. Pick a category (Base, Fine-Tune, Quant) or leave on All for a complete view.
2. Apply optional filters for family, license, access, and timeframe to focus your search.
3. Switch between downloads and likes sorting, and adjust the Max Results slider to control the table size.
4. Toggle Show Hidden Gems only to highlight promising lesser-known models.
5. Click Refresh Data whenever you want to clear the cache and pull brand-new stats.

Developed by Tyler Williams