Bili Note ingests online or local videos (Bilibili, YouTube, Douyin, etc.) to transcribe and summarize them into structured Markdown with screenshots and jump links. It supports GPT and local Fast-Whisper for audio and summarization.
Users can pick note styles, keep multiple versions, and self-host with FastAPI backend plus Vite/React frontend. FFmpeg is required; GPU acceleration via CUDA is optional for faster transcription.
Use Cases:
BiliNote v1.8.1
AI 视频笔记生成工具 让 AI 为你的视频做笔记
BiliNote 是一个开源的 AI 视频笔记助手,支持通过哔哩哔哩、YouTube、抖音等视频链接,自动提取内容并生成结构清晰、重点明确的 Markdown 格式笔记。支持插入截图、原片跳转等功能。
详细文档可以查看这里 (https://docs.bilinote.app/)
可以通过访问 这里 (https://www.bilinote.app/) 进行体验,速度略慢,不支持长视频。
本项目提供了 Windows 系统的 exe 文件,可在release (https://github.com/JefferyHcool/BiliNote/releases/tag/v1.1.1)进行下载。**注意一定要在没有中文路径的环境下运行。**
!screenshot !screenshot !screenshot !screenshot !screenshot
git clone https://github.com/JefferyHcool/BiliNote.git
cd BiliNote
mv .env.example .env
cd backend
pip install -r requirements.txt
python main.py
cd BillNote_frontend
pnpm install
pnpm dev
访问:http://localhost:5173
本项目依赖 ffmpeg 用于音频处理与转码,必须安装:
# Mac (brew)
brew install ffmpeg
# Ubuntu / Debian
sudo apt install ffmpeg
# Windows
# 请从官网下载安装:https://ffmpeg.org/download.html
⚠️ 若系统无法识别 ffmpeg,请将其加入系统环境变量 PATH
若你希望更快地执行音频转写任务,可使用具备 NVIDIA GPU 的机器,并启用 fast-whisper + CUDA 加速版本:
具体 fast-whisper 配置方法,请参考:fast-whisper 项目地址 (http://github.com/SYSTRAN/faster-whisper#requirements)
确保你已安装 Docker 和 Docker Compose:
docker 部署 (https://github.com/JefferyHcool/bilinote-deploy/blob/master/README.md)
抖音下载功能 部分代码参考引用自:Evil0ctal/Douyin_TikTok_Download_API (https://github.com/Evil0ctal/Douyin_TikTok_Download_API)MIT License
💬 你的支持与反馈是我持续优化的动力!欢迎 PR、提 issue、Star ⭐️
如果你觉得项目对你有帮助,考虑支持我一下吧
Star History Chart (https://api.star-history.com/svg?repos=JefferyHcool/BiliNote&type=Date)
Real-time desktop overlay copilot that watches your screen and listens to calls, delivering contextual answers with profile presets for interviews, sales, and presentations.
Proactive context-aware desktop assistant that captures your screen, stores data locally, and uses OpenAI/Doubao-compatible models to deliver summaries, insights, and todos.
Markdown-native project board for Git repos with CLI, web UI, MCP/AI integration, and offline-friendly Kanban/search built entirely on plain Markdown files.
Peer-to-peer, end-to-end encrypted file and folder transfer over QUIC with resumable downloads, no accounts, and cross-platform desktop builds.
Context-aware Windows overlay assistant that reads your screen and delivers translations, summaries, and answers via multi-LLM backends with a sleek keyboard-driven UI.
WebRTC P2P tool for files, text, and desktop sharing with end-to-end encryption, ACK reliability, Docker/single-binary deploys, and a responsive Next.js UI.