Gemini Embedding 001 — HYR Studio's Semantic Memory
Search your whole back-catalogue by meaning. Get answers grounded in your own content.
Gemini Embedding 001 converts every source, episode, and library item into vectors so HYR Studio can search your entire content library by meaning — and ground every Notebook Chat answer in your own material.
Provider
Google DeepMind
How HYR Studio uses Gemini Embedding 001
Everything you import is embedded on ingest. Those vectors power semantic search across sources, podcast episodes, and library items, Episode Routing content matching, and retrieval-augmented answers in Notebook Chat.
Capabilities
- High-quality text embeddings
- Semantic search across your full library
- Retrieval-augmented generation (RAG) grounding
- Cross-episode topic matching
Output Types Powered
- Semantic search results
- Grounded Notebook Chat answers
- Related-content matching
This is why a 1,000-video back-catalogue becomes searchable in an afternoon: every transcript is findable by what it means, not just the words it contains.