Memories.ai gives Claude long-term visual memory. Connect your memories.ai account and Claude can search, recall, and reason over your own indexed video library — by what's seen (visual scenes), said (transcripts), or described (captions), not just by filename. With Memories.ai, Claude can: • Index any public video URL into your private library and poll until it's searchable. • Semantically search your library across clip, audio, image, and caption modalities — or run exact-text search over transcripts. • Pull scene-by-scene visual captions and full spoken-word transcripts for any indexed video. • Search the live public web (YouTube, TikTok, Instagram, X) in real time, or scrape one social video's metadata or transcript by URL. • Ask a vision-language model free-form questions about any video or image URL, and transcribe audio to timestamped text. • Manage the library — list videos, check indexing status, and delete videos. You bring your own memories.ai account: connect once via OAuth and every request runs against your data with your credentials. The connector forwards your key per request and stores nothing itself. Ideal for creators, researchers, and teams turning a growing video archive into searchable, question-answerable memory.
Anthropic’s own published signals, snapshot of August 23, 2026; units undocumented — treat as ordinal.
Over time
Details
Tools(16)
- analyze_image
- analyze_video
- delete_videos
- get_audio_transcription
- get_video_caption
- get_video_metadata
- list_unique_ids
- list_videos
- scrape_social_video
- search_audio_transcripts
- search_public_videos
- search_videos
- search_web_videos
- transcribe_audio
- upload_video_from_file
- upload_video_from_url
Data from Claude’s public directory API, snapshotted daily by the MCP App Tracker. A listing describes what the directory publishes — not an endorsement.