I made a tool to search your video footage and it's on GitHub
Local YOLO and Whisper indexing beats Google's $400 API bill for private video search.
Local-first Video Knowledge Base. Index your video library with multi-modal analysis (YOLO, DeepFace, Whisper), search semantically via natural language, Docker-ready.
Replaces a $400 Google API bill with local YOLO, DeepFace, and Whisper running in Docker.
Video editors, content creators, and privacy-conscious archivists
Google Video Intelligence API · Twelve Labs · VideoDB
I tried the Google Video Intelligence API. Got a $400 bill for 4 videos (5 minutes average, 4k videos) of analysis (doesn't include video transcription), and I used my GCP startup credits to cover the bill.
I decided to build my own tool that needs to have 3 important things: can transcribe videos, analyse video frames, and everything needs to be done locally.
I don't wanna deal with storing my videos in the cloud because of two concerns: privacy and storage cost.
I've been working for the last couple of months. I have a source available version that can be used for free (personal and commercial use with companies that have fewer than 5 people). Available here (https://github.com/IliasHad/edit-mind), and the project has 1.3k Github stars
Now, I'm building a desktop app with direct NLE integration (Final Cut Pro, DaVinci Resolve, and Adobe Premiere Pro). This includes an editing agent that understands your footage and your editing style.
Preview: https://youtu.be/jcctyfVg_34
Happy to answer questions and hear your feedback.
Local YOLO and Whisper indexing beats Google's $400 API bill for private video search.
Zero-knowledge encryption for transcripts is nice, but Otter.ai and TurboScribe exist.
Metal sphere rendering with GPU color conversion, but panorama players already exist.
Bilibili-specific RAG pipeline with fallback ASR for inaccessible audio URLs.
MCP integration with Cursor and Claude Code sets this apart from generic RAG tools.
Yet another local video indexer competing with Kive, Frame.io, and Adobe's built-in search.