model-cards — Browse and Compare Open-Weight AI Models from Your Terminal

There are now 20+ viable open-weight models competing for your attention — Qwen 3.8, DeepSeek V4, Kimi K2.6, GLM-5.2, MiniMax M3, and more releasing every week. Keeping track of who has what context window, which license each uses, and how much hardware you need to run them is a full-time research job.

Today I’m releasing model-cards — a CLI reference tool that queries Hugging Face’s model card API and displays specs in your terminal. Search, browse, and compare models without leaving the command line.

What it does

# Search for a model family
model-cards search qwen

# View full details
model-cards show deepseek-v4

# Compare side-by-side
model-cards compare qwen3-coder kimi-k2.6 deepseek-v4

Features

• Search — find models by name, description, or task type (sorted by downloads)
• View details — full model card: benchmarks, hardware requirements, license, context window, architecture, release date
• Compare — side-by-side comparison of up to 4 models. Highlights the best value in each spec category
• Live data — always fresh from Hugging Face API, no bundled JSON to go stale
• Rich terminal output — color-coded tables, panels, and formatted numbers
• JSON output — --json flag for pipeline-friendly consumption

Why this matters

With the open-weight model explosion, the bottleneck isn’t model quality anymore — it’s discovery. Which model has 128K context? Which one is MIT-licensed? Which one runs on a 24GB GPU? Model-cards answers these questions in seconds.

Try it

pip install model-cards
# or clone from GitHub
git clone https://github.com/vikasudasi/model-cards
cd model-cards && pip install -e .

GitHub →