Free RAG app stack
Answer questions over your own documents: an LLM, embeddings in a vector store, an API host and a bucket for the files.
| Role | Pick | What you get | Swap for |
|---|---|---|---|
LLM API Generates the answer from retrieved passages. | LLM API  | Flash models, free of charge | |
Vector store Holds embeddings and finds the nearest passages. | Vector store  | 1 GB RAM cluster, forever | |
API hosting Runs the retrieval and prompt code. | API hosting  | 100,000 requests a day | |
Document storage Keeps the source files you index. | Document storage | 10 GB, zero egress fees |
Price: Free of charge input and output (Flash) · Rate limits: RPM, TPM and RPD per project · Reset: Daily quota resets at midnight Pacific · Data use: Free-tier content used to improve Google products · Batch, Flex: Not available on free tier
Sign upCluster: Single node · Size: 0.5 vCPU, 1 GB RAM, 4 GB disk · Inference: Free on selected models
Sign upRequests: 100,000 / day · CPU time: 10 ms per invocation · KV reads: 100,000 / day · KV writes: 1,000 / day · KV storage: 1 GB
Sign upStorage: 10 GB-month / month · Class A ops: 1 million / month · Class B ops: 10 million / month · Egress: Free
Sign upOffers change often; the vendor's page is the final word. Links go straight to the vendor, no affiliate links. Logos via logo.dev; trademarks belong to their owners.