How do I save my first item?

Paste any YouTube URL, article, or PDF into the capture bar. BrainTube extracts the transcript, generates a summary, and indexes it automatically.

How do I connect BrainTube to Claude or Cursor?

Open /connect in the app to get your MCP URL and API key, then paste it into Claude Desktop's config or Cursor's MCP settings. Your library becomes a tool the AI can call.

A Brain is a focused group of items you can chat with. Dashboard → Brains → New, pick items, give it a name. You can keep it private or share it publicly via a slug.

How does billing work?

Plans run monthly or annually, with a 14-day free trial on Starter and Pro and no card needed. Each plan includes monthly credits for AI actions. Refunds are handled case by case.

Where is my data stored?

Your library is stored in the EU, private to your account. AI providers process content to transcribe, summarise and embed it. You own your data: export your library as Markdown or JSON from Settings, or the whole corpus through MCP. Account deletion purges backups within 30 days.

Available MCP Tools — Help

The main tools on BrainTube's MCP server, by group. Every account gets 38 tools; the live list is on the Connect page.

Every BrainTube account gets 38 MCP tools, 10 of them write-capable: save content, bulk save, add notes, tag, bookmark, compile knowledge, and Notion and Readwise import. MCP access and bt_ API keys are included on every BrainTube plan, including Free. The live list, with each tool's own description, is on the Connect page.

The main tools, by group:

Search and retrieval

search_knowledge: hybrid semantic and keyword search across your library
deep_search: follows the knowledge graph from the top results (multi-hop)
search_by_source and search_by_date_range: narrow by source type or date
get_related and find_path: explore how items connect
get_video and list_recent: read one saved item, or see your latest saves

Saving and organising

ingest_content: save one URL or piece of content
bulk_ingest: save many items in one call, with deduplication
add_note, tag_item and toggle_bookmark
compile_knowledge
ingest_notion_page, ingest_notion_database, connect_readwise and sync_readwise

Knowledge graph

get_knowledge_graph, get_knowledge_index and get_concept_articles
entity_cooccurrence, tag_cooccurrence, detect_gaps and knowledge_health

Review and context

random_resurface: bring saved items back for review
get_session_brief, get_recent_conversations and get_expertise_profile

Brains and export

list_brains and chat_with_brain
export_corpus: your whole library as a ZIP of Markdown files, returned as a signed download URL
export_claude_md

Best practices

Start with search_knowledge for broad questions
Use get_related to explore connections from search results
Chain search_knowledge and get_knowledge_graph for deeper exploration
Let your AI save what it finds with ingest_content or bulk_ingest