ChannelAI — Open Source Channel Automation
Free MIT-licensed YouTube channel automation — in-browser reel rendering, caption export, thumbnail capture and scheduled publishing.
ChannelAI is a fully open-source web app that automates the repetitive parts of running a video channel: planning clips, splitting long footage into reels, rendering them for real in the browser, writing caption files, generating titles and hashtags from templates, and publishing to YouTube on a schedule. Every render happens client-side — canvas crop, MediaRecorder and WebAudio produce a genuine WebM/MP4 file that the user downloads. Files never leave the device; the only account involved is the user’s own free Google OAuth client against the official YouTube Data API v3. Scheduled publishing is enforced by YouTube itself through `publishAt`, so an upload can be scheduled and the tab closed.
Key Features
- Real in-browser clip rendering — canvas crop, re-encode, progress and cancel
- Local auto-split that turns one footage range into 2–6 equal reel clips
- Timed caption line editor with genuine SRT and VTT export
- Real frame capture for thumbnails with full-resolution PNG download
- Local `{{variable}}` template engine for titles, descriptions and hashtags — no AI metering
- Scheduled YouTube publishing through the official Data API using the user’s own OAuth client
- In-app and system notifications for daily posting cadence rules
Repository Topics
Challenges
Doing real video encoding, caption generation and API publishing entirely client-side, while keeping the honest boundary that no static web app can fetch another site’s video bytes.
Results
A genuinely free, MIT-licensed, self-hostable alternative to paid credit-metered tools — unlimited channels, unlimited usage, and no upload of user footage.