Pinecone through Squiid
Managed vector database for retrieval augmented generation at scale. Pinecone is next up on Squiid, being built now.
What Pinecone is
Pinecone is a vector database built as a managed service. You create an index with a dimension and a distance metric, upsert vectors with metadata attached, and query for the nearest neighbours to an embedding, optionally filtered by that metadata. Serverless indexes decouple storage from compute so you pay for what you read and write rather than for a running pod, namespaces partition one index per tenant or per document set, and integrated inference can generate the embeddings for you so a record goes in as text. Hybrid search combines dense vectors with sparse keyword matching for queries where an exact term matters.
In a vibe-coded app this is the memory layer. Product documentation that a support bot answers from, a company knowledge base, transcripts of past calls, a user's own uploaded files, or an agent's long-term memory all get chunked, embedded and written here, then retrieved by similarity at question time. Metadata filtering is what keeps it correct in practice, because a query has to be restricted to the right customer, the right document version or the right date range before relevance is even considered. Rebuilding an index is also a normal event, because changing the embedding model means every vector has to be regenerated.
Why connect it through Squiid
Squiid provisions the Pinecone project and holds the API key, so a retrieval service that is often called from many small functions does not need the credential copied into each one. Requests go to https://api.squiid.io/v1/pinecone/... with Authorization: Bearer $SQUIID_API_KEY, and the official clients work by pointing their base URL at the gateway (gateway convention). Reads, writes and storage are metered against prepaid credits at Pinecone's list price, and an index that is quietly growing shows up in the dashboard beside the model spend that fills it.
- One account, one key, one bill, one dashboard. Pinecone sits beside every other service your project uses, on the same invoice and the same credit balance.
- Key custody. Squiid holds the Pinecone credentials. Your code carries
SQUIID_API_KEY, so PINECONE_API_KEY, PINECONE_INDEX, PINECONE_ENVIRONMENT never land in a repository, a preview deployment or an agent’s context. - Metering you can see. Every call is counted per service and per day, so you know what Pinecone costs before the month ends.
- Pause on zero. When prepaid credits run out the gateway stops authorizing usage instead of running up a bill you did not fund.
- One rotation. Rotating the Squiid key cuts off every agent, worker and deployment at once, with no provider key to hunt down.
What you get
- A provisioned Pinecone project with the key held by Squiid
- Serverless indexes with metadata filtering and namespaces
- Hybrid dense and sparse search
- Reads, writes and storage metered against credits
- Pinecone list pricing, no markup
How to connect Pinecone
One variable replaces the provider credentials this service would normally need. Requests go to the gateway and the official SDKs work by pointing their base URL at it (gateway convention).
# .env
# replaces PINECONE_API_KEY, PINECONE_INDEX, PINECONE_ENVIRONMENT
SQUIID_API_KEY=sq_live_xxxxxxxxxxxxxxxxxxxxxxxx
# every request: https://api.squiid.io/v1/pinecone/...
# header: Authorization: Bearer $SQUIID_API_KEY
Add it in your dashboard
Open the Squiid catalog, add Pinecone and fund credits. Squiid provisions the project and keeps the provider keys.
Put the one key in your env
Add SQUIID_API_KEY to .env and delete PINECONE_API_KEY, PINECONE_INDEX, PINECONE_ENVIRONMENT. Nothing else from Pinecone needs to exist in your project.
Point your agent at it
Claude Code, Cursor and Codex read the same variable and call https://api.squiid.io/v1/pinecone/. Claude Code and Cursor usually write the chunk, embed and upsert pipeline as a background job, then query Pinecone inside the chat route before building the prompt.
Pricing through Squiid
Pinecone charges for read and write units plus stored vector data on serverless indexes, with plan tiers above the free one, at the provider's listed price. Squiid passes that through at par: one dollar of prepaid credit is one dollar of Pinecone usage, with no markup and no per-request margin.
Pinecone has no subscription attached on Squiid today. If a plan is required for a feature you need, it passes through at the provider’s listed price with no Squiid fee.
Squiid adds no markup to provider prices. There is a free plan at $0 a month with a 15% fee when you top up credits, Solo at $19.97 a month with a top-up fee that starts at 8% and falls to 3.5% as your spend grows, Team from $29.97 a seat a month with the same ladder one point lower, and Custom with a rate of 3% or less. Credits are prepaid, so Pinecone usage stops when the balance reaches zero rather than producing an invoice you did not fund. Auto top-up with a cap keeps production running without handing anyone an open-ended card.
If you leave
Export: vectors and index settings are exported and re-imported into an index in your own Pinecone account. Squiid never owns your code, your schema or your domain, and the gateway speaks Pinecone's own API, so the client change is a base URL. The handover page lists what happens for every provider in the catalogue.
Works with every coding agent
Pinecone through Squiid is just an HTTP call with a bearer token, so every agentic coding tool can use it with the single key already in your environment:
- Claude Code
- Cursor
- ChatGPT
- Codex
- Grok
- Gemini CLI
- GitHub Copilot
- Windsurf
- Perplexity
- Devin
- Replit
- v0
- Lovable
- Cline
Vibe coding a feature on Pinecone usually means asking for it in plain language and letting the agent write the integration. Because the key is Squiid’s rather than the provider’s, it can do that without ever holding a credential that could be leaked, logged or committed.
Related services
Qdrant Cloud
Vector and searchOpen-source vector search with rich payload filtering and quantization
RoadmapProvisioned WEWeaviate
Vector and searchVector database with built-in vectorization modules and hybrid search
RoadmapProvisionedOpenAI
LLM APIsGPT models, embeddings, images and realtime voice over one API
LiveInstant COCohere
LLM APIsEnterprise models with strong embeddings, rerank and RAG tooling
RoadmapInstantSupabase
DatabasesPostgres with auth, storage, realtime and edge functions attached
Next upProvisionedQuestions people ask
Can Claude Code build a RAG pipeline on Pinecone through Squiid?
Yes. The embedding provider and Pinecone are both behind the same SQUIID_API_KEY, so the agent writes one credential into the environment and nothing else.
Does Squiid mark up Pinecone usage?
No. Read units, write units and storage bill at Pinecone's list price against your prepaid credits.
Pinecone or a vector column in Postgres?
pgvector in Supabase or Neon is simpler when the corpus is small and already in your database. Pinecone is built for large indexes, high query rates and tenant namespaces.
How do I keep one customer's documents out of another's results?
Use a namespace per tenant, or a metadata filter applied on every query. Both work unchanged through the gateway.
What happens if credits run out?
Queries and upserts stop being authorized. Your index and its vectors are not deleted for a zero balance.
Connect Pinecone with one key.
Pinecone stores embeddings and returns the nearest matches in milliseconds, with metadata filtering, namespaces for multi-tenancy and serverless indexes that scale to zero. Provisioned by Squiid on one key.