Chroma InvalidDimensionException: Embedding Dimension Does Not Match Collection

How to fix it

Try these in order — the most common fix is first.

1. 1. Use the matching embedding model

   Copy steps
   1. Check which model created the collection
   2. Pass the same embedding_function on every get_collection call
   Did this fix it? Yes No

2. 2. Start a new collection for the new model

   Copy steps
   1. Create a new collection
   2. Re-embed and add your documents
   3. Delete the old collection once verified
   Did this fix it? Yes No

Symptoms

InvalidDimensionException: Embedding dimension 1536 does not match collection dimensionality 384 (numbers vary).

Why it happens

• Switched embedding model after the collection was created
• Collection reopened with a different embedding function
• Upgrading to 0.5.0 and opening old persistent data (per the Chroma cookbook)

This guide also covers

• chromadb embedding dimension does not match
• InvalidDimensionException chroma
• chroma dimension 384 1536

Other Chroma errors

• Chroma "Your system has an unsupported version of sqlite3" Error
• Chroma pip install "Failed to build hnswlib" Error
• Chroma Data Disappears After Restart (Persistence Fix)
• Chroma "Collection Not Found" Error
• Chroma "Collection Already Exists" UniqueConstraintError
• Chroma "sqlite3.OperationalError: database is locked"
• Chroma "Could not connect to tenant default_tenant" Error
• Chroma "Cannot return the results in a contiguous 2D array. Probably ef or M is too small"