Models & local AI
Open models, and the tools to run them on your own hardware. 22 repos, most starred first.
transformers
Transformers is the Hugging Face Python library that defines pretrained models for text, vision, audio and multimodal tasks. Pick it for one consistent API to run or fine-tune models from the Hugging Face Hub.
#2open-webui
Self-hosted chat interface for Ollama and OpenAI-compatible models, with RAG and plugins
#3odysseus
Self-hosted workspace for AI chat, agents, research, documents, email and calendar
#4minimind
Train a tiny 64M-parameter language model from scratch on one consumer GPU
#5Kronos
Pre-trained Transformer models that forecast candlestick (K-line) price data
#6airllm
Run very large open language models on a small GPU by loading one layer at a time
#7Vane
Self-hosted AI answer engine that searches the web and cites its sources
#8timesfm
Pretrained Google model that forecasts time series without training on your data
#9Unlimited-OCR
Open OCR model that turns images and multi-page PDFs into structured text in one pass
#10gallery
Android and iOS app for running open LLMs offline on your phone
#11lingbot-map
Feed-forward model that builds 3D scenes from video frames as they stream in
#12needle
A tiny 8-29 MB model that calls tools and extracts fields on phones and microcontrollers
#13cosmos
Open world models that reason about and generate video, sound and robot actions
#14pinokio
A desktop launcher that installs and runs open-source AI projects with one click
#15osaurus
A native macOS app for running AI agents with local or cloud models
#16ODS
Installer that sets up a private local AI server with chat, voice, RAG and workflows
#17GLM-5
Open-weights LLM family from Z.ai built for coding and long-running agent tasks
#18GLiNER
Small zero-shot model that pulls any entity types out of text, CPU-friendly
#19tribev2
Predicts fMRI brain responses to video, audio and text, with pretrained weights
#20HyperQwen
vLLM patches and Docker setups to run Qwen3.8-27B fast on one 24 GB GPU
#21StartLux-Decision
Open models that give a probability for every answer option, and think only when unsure
#22claudish
Two tiny local models that translate plain English to and from over-engineered prose