Pinecone Vector Dimension Does Not Match the Dimension of the Index

How to fix it

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

1. 1. Match the index to the model

   Copy steps
   1. Check your embedding model's output size in its docs
   2. Create a new index with that exact dimension and the right metric
   3. Re-embed and upsert your data into the new index
   Did this fix it? Yes No

2. 2. Fix the producer instead

   Copy steps
   1. Print len(vector) before upserting
   2. Set the model's dimensions parameter, if it supports one, to the index size
   3. Use the same model for documents and queries
   Did this fix it? Yes No

Symptoms

Upsert or query fails with a 400 error such as 'Vector dimension 768 does not match the dimension of the index 1536'.

Why it happens

• Embedding model output length differs from the index dimension (for example 768 vs 1536)
• You recreated the index with a new dimension but queries still use the old model
• Wrong dimension setting in a no-code node or library (values like 0, 1 or 8 point to a config bug)
• Documents and queries embedded with different models

This guide also covers

• vector dimension does not match the dimension of the index
• pinecone dimension mismatch
• pinecone 400 bad request upsert

Other Pinecone errors

• Pinecone 429 Too Many Requests Error
• Pinecone Upsert Failing: Batch and Size Limits
• Pinecone Query Returns No Results or Empty Matches
• Pinecone Upsert Not Showing Up After Writing
• Pinecone Starter Index Paused, Archived or Deleted for Inactivity
• Pinecone Starter Plan Limits Reached
• Why Is My Pinecone Bill So High
• How to Cancel Pinecone or Downgrade to Starter