Open source AI, by area

Autonomous agents and the frameworks to build them — orchestration, planning, multi-agent systems. Tracked daily with star velocity from real snapshots.

The Model Context Protocol ecosystem: servers, clients and tooling that let agents talk to the outside world.

Agents that write, review and ship code — from terminal copilots to full autonomous engineers.

Model-layer infrastructure: training, fine-tuning, prompting and everything that wraps a large language model.

Retrieval-augmented generation: pipelines, embeddings and search layers that ground models in real data.

Running models fast and cheap: inference engines, serving stacks, quantization and edge deployment.

Where agents remember: vector stores, memory layers and similarity search engines.

Agents that drive real interfaces — browser automation, computer use and web agents.

Visual builders and engines that chain models, tools and triggers into working automations.