Discover
Every project in the index. Underrated first: good projects that haven’t been noticed yet.
- 1.5k
Harvey LAB
Benchmark tasks and harness for testing AI agents on legal work
- 1.9k
Video Search and Summarization
Reference architecture for AI agents that search and summarize video
- Linzhan Mou2.1k
UniMate
Text-driven motion generation model for characters with any skeleton
- 2.3k
stable-worldmodel
Python toolkit for collecting data, training, and evaluating world models
- 2.2k
SIA
Self-improvement loop that rewrites an AI agent to score higher on a benchmark
- 2.5k
Microduck RL
Reinforcement learning environments for a small bipedal robot
- 2.6k
Stable Video Infinity
Research code and weights for generating very long videos without drift
- 2.7k
TokenSpeed
LLM serving engine tuned for agentic workloads on datacenter GPUs
3.2k
- Antoine2.3k
forge
Reliability layer that makes local LLM tool calls succeed more often
- 3.4k
SIE
Self-hosted inference server running many open models behind one API
AI models and runtimesAgent frameworks and toolsServers and infrastructure
- 4k
trackers
Python library of multi-object tracking algorithms for any detector
- 4k
Marin
Open research platform for training large language models end to end
- 4k
PersonaLive
Real-time diffusion model that animates a portrait photo from live video
- 4.2k
MOSS-TTS
Open speech and sound model family for TTS, dialogue, and effects
- Maziyar Panahi5.7k
OpenMed
Clinical entity extraction and PII de-identification on local hardware
AI models and runtimesHealth and fitnessPrivacy and security
- az5.7k
RLM
Inference library letting language models recursively call themselves via a REPL
- 5.7k
UltraRAG
MCP-based framework for building and evaluating RAG pipelines in YAML
- 5.8k
Newton
GPU-accelerated, differentiable physics engine for robotics research
- 5.9k
AReaL
Asynchronous reinforcement learning system for training LLM agents
6.2k
Prince Canuma5.6k
- 7.3k
vllm-omni
Serving engine for multimodal, diffusion, speech, and action models
- Keller Jordan6k
Modded-NanoGPT
Community speedrun for training a GPT-2-scale model as fast as possible
