RIP, vector database
Turbopuffer is transitioning its storage architecture to a new engine, informally called 'v3', to move beyond its reliance on ANN vector indexes. This change aims to allow the database to handle more complex query plans and scale more effectively.
Why it matters
This shift reflects a broader trend in database engineering where specialized vector search tools are evolving into more generalized, high-performance search engines.
tpuf v3 is coming: follow along turbopuffer v3 is coming: follow our progress Navigation turbopuffer Use Cases Docs Pricing Customers Blog Jobs Contact Log in Sign up RIP, vector database September 30, 2026 • Dan Harrison (Engineer) We are changing turbopuffer's storage architecture. The new engine, informally called "turbopuffer v3", changes how documents and indexes are laid out, written, compacted, and queried in turbopuffer. It will allow us to serve many more query plans, at much greater scale.
turbopuffer launched as a serverless vector database , highly specialized to the task of serving extremely cheap and reasonably fast vector searches. Object storage as the source of truth gave the economics, and tiered NVMe SSD/memory caches gave the performance. The value of these particular tradeoffs was validated by our earliest customers, including Cursor and Notion.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in