Claude Video
A free tool that lets Claude actually "watch" a video by pulling out the key frames and a timestamped transcript, instead of only being able to read a video's title and description.
🔗 Visit Claude Video
Description
Claude can't watch a video directly the way a person can, so understanding what happens in a YouTube video, a TikTok clip, or a local video file normally requires someone to describe it manually. Claude Video bridges that gap: it extracts the important frames from a video (skipping redundant ones to save tokens) and generates a timestamped transcript, giving Claude enough visual and textual context to actually analyze what's happening.
Claude Video is a free, open-source (MIT license, ~10,000 GitHub stars, 1,100+ forks) Python tool that extracts video frames at configurable detail levels and generates timestamped transcripts from existing captions or via Whisper transcription, with intelligent frame deduplication to reduce token usage. It supports YouTube, TikTok, Vimeo and local video files, offers multiple analysis modes (transcript-only, efficient, balanced, and a more thorough "token-burner" mode), and supports time-range-focused analysis for looking at just a specific portion of a longer video. It requires ffmpeg and yt-dlp for video processing, with optional Whisper API keys (Groq or OpenAI) for transcription.
💬 Our review
The short version: if you want Claude to genuinely analyze video content rather than just read a title, Claude Video's frame-extraction-plus-transcript approach with smart deduplication is a well-built, free way to do that efficiently.
With over 1,100 forks against 10,000 stars, the fork ratio here is unusually high, suggesting this individual developer's project is being actively adapted and extended by real users rather than just bookmarked — a healthier adoption signal than star count alone. The multiple analysis modes (from token-efficient to thorough) is a thoughtful design choice that lets you trade off cost against depth of understanding depending on the task, rather than forcing one fixed approach. The setup does require some technical comfort (ffmpeg, yt-dlp, optional Whisper API keys), so it's aimed at developers integrating video analysis into their own Claude-based workflows, not a plug-and-play consumer app — but for that audience, it's a solid, actively-used free tool.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Entièrement gratuit et open source (licence MIT). Clés API Whisper (Groq ou OpenAI) optionnelles pour la transcription, à la charge de l'utilisateur.
Pros
Gratuit, open source, adoption réelle prouvée (1 100+ forks)
Déduplication intelligente des frames pour économiser des tokens
Plusieurs modes d'analyse selon le compromis coût/profondeur souhaité
Supporte YouTube, TikTok, Vimeo et fichiers locaux
Cons
Nécessite ffmpeg et yt-dlp installés, barrière technique réelle
Clés API Whisper optionnelles à payer séparément pour la transcription
Projet individuel, pas de site dédié ni de support commercial
Pertinent seulement pour des workflows développeurs autour de Claude