本地大模型运行时、AI 编程助手、推理引擎与前沿 RAG 架构。
Distill Jev into a local model, with a disagreement bound
We compressed a 31B video+audio diffusion model to 14B, 6 steps, INT8
I built a tool to create AI agents that live in iMessage
math app for kids where the AI is wrong on purpose
Cut audio by selecting text, fully on-device
Turn Any Photo into an AI Meme Video (No Prompts Needed)
open-source tab-autocomplete
an open schema for defining AI agent teams
A 40 MB GPU cluster manager for vLLM/SGLang, no K8s needed