## TL;DR

Vectors from different models cannot be compared in the same index: a 384-dim local model and a 1024-dim hosted model need separate Atlas vector indexes.

## Fix

1. Vectors from different models cannot be compared in the same index: a 384-dim local model and a 1024-dim hosted model need separate Atlas vector indexes.
   Expected: you can confirm this matches your situation.

## When to use

- You are setting up or using this MongoDB Atlas Vector Search feature.
- The symptom matches: dimension mismatch after switching embedding models.

## When NOT to use

- Unrelated MongoDB Atlas Vector Search issues (different feature, different failure).
- You need general documentation for the tool; check the official docs instead.

## Compatibility

Reported against MongoDB Atlas Vector Search. Source: a local/non-public address.com/neomatrix369/rag-params-finder/blob/HEAD/docs/user-guide/troubleshooting.md.

## Variant phrasings

### dimension mismatch after switching embedding models

## Why it happens

Vector search fails with a dimension error, or results are nonsensical, after switching the embedding model.

## Edge cases

- If your error message differs even slightly, this is probably a different issue; search the exact text.
- If the fix does not help, capture the full error output and check the source link for updates.

## Source

a local/non-public address.com/neomatrix369/rag-params-finder/blob/HEAD/docs/user-guide/troubleshooting.md