**TL;DR:** Use Upstash Vector for embeddings and Redis for everything else: sessions, counters, rate limits, and metadata live in Redis. Query by meaning, filter by metadata. - Upstash Redis: everything with a TTL, a counter, or exact-key access. Add Vector when recall needs semantic search over large fact stores instead of full-text search.

## The fix

1. Upstash Vector: embeddings and nearest-neighbor search. Query by meaning, filter by metadata.

2. Upstash Redis: everything with a TTL, a counter, or exact-key access. Sessions, rate limits, caches, queues, feature flags, chat history. Do not store embeddings in Redis and brute-force similarity. Do not store session TTLs in Vector. Each product is built for its half.

3. Vector index holds document chunks with metadata { docId }.

4. Redis holds doc:DOCID with the full text, title, and TTL. Retrieval: vector search returns chunk IDs, then MGET the Redis keys for the full payloads. Vectors stay small, Redis serves the bytes.

## When to use this

- This covers exactly what the title says: Upstash Vector plus Redis.
- You are setting this up for the first time, or auditing an existing setup.
- You want the key gotchas in one place before you start.

## When not to use this

- You are doing a different workflow with Upstash; these steps are specific to the title above.
- You need the full reference docs; this is the short path, not the manual.

## Compatibility

- Not pinned to a specific version; follows current Upstash behavior.