Home
DESIGN.mdStar on GitHub
The vector database

Give agents knowledge

The vector database to build knowledgeable AI — store, search, and retrieve billions of embeddings with low latency.

Brand
Brand Blue
#002BFF
Actions, high sim
Periwinkle
#5B7AEE
Support, med sim
Violet
#3D11A0
Deep accent
Frost Blue
#B3D5FF
Tints, fills
Accent spectrum
Purple
#A440CE
Cyan
#8BF6FF
Blue-violet
#1C17FF
Success
#15B077
Surfaces
Base
#FBFBFC
Light page
Surface
#F2F3F6
Cards
Dark base
#171717
Neutral, not navy
Error
#E31957
Destructive
display · GT Planar
Give agents knowledge
GT Planar500
h2
Build knowledgeable AI
22px500
body
Store, search, and retrieve billions of embeddings with low latency.
15px400
pixel · Apple II
SEARCH COMPLETE — 5 MATCHES
Apple II
mono
index.query(vector=embed(q), top_k=5)
JetBrains Mono
Buttons
Similarity band
≥ 0.85 high 0.70–0.85 < 0.70 low
knowledge-base · 1536-d metric cosinevectors 2,481,004pods p1.x1
reset password pricing upsert docs
Press Search to query the index
Ready · top_k = 5 · namespace default

Query the index → top-K nearest vectors ranked by cosine similarity, colored by band.

Unofficial design reference · built from live pinecone.io · not affiliated with Pinecone