[NeurIPS 2025] Radial Attention: O(nlogn) Sparse Attention with Energy Decay for Long Video Generation

做企业 AI 视频项目逼出来的工具包 · 411 prompt · 15 模型 · 7 Claude Skill · 14 篇方法论

Mochi is a proof-of-concept C++ loader that leverages the ChaiScript embedded scripting language to execute code.

A hybrid kernel that simply refuses to die. While other OSs are busy collecting Blue Screens, mochiOS is busy actually running your code

ESP32 client for the GenAI Mochi animatronics project

Based on Dasai™ Mochi

Create flashcards on Mochi.cards using the API provided by Mochi's Pro subscription in Obsidian.

Go server for the GenAI Mochi animatronics project

Like Anki's Review Heatmap, but for Mochi Cards!

Convert your Notion database to Mochi cards.

Dataset preparation tool for Mochi-1 video model fine-tuning — clip extraction, captioning & formatting pipeline

WhatsAI - An easy-to-use UI fully based on ComfyUI.

Serve Text-to-Video Models in Production

Front-end framework for developing responsive, mobile-first projects on the web or offline.

Generate videos with open-source AI models (Mochi). Built as a free tool for contestants of the Project Odyssey 2024 AI Film Competition (projectodyssey.ai). Experiment — archived.

Benchmarking training throughput of contemporary diffusion models on AMD's GPU hardware.

Generate videos with Mochi as an endpoint on Runpod

One-way synchronization of flashcards from Logseq to Mochi

A high-performance, GPU-optimized real-time speech-to-text (STT) streaming server built with WebSocket support for multiple concurrent clients. This project leverages the Kyutai STT model and is optimized for NVIDIA RTX 4090 GPUs, providing low-latency transcription for audio streams.

A pen-and-paper based, efficient and scalable spaced repetition system.