Pinecone diagnosis: upsert rejected, vector dimension vs index dimension

Diagnosis

Symptom

Upsert (or query) fails with a dimension-mismatch message: the vector has N dimensions, the index expects M.

Likely cause

The embedding model producing the vectors is not the model the index was created for. Index dimension is frozen at creation; the model is the moving part.

Confirm it

1. Print len() of one embedding straight from your model.
2. Run pc.describe_index(name) and read dimension.
3. Check whether the model changed recently (dependency bump, config edit) or the index was created for a different model.

Fix

1. If the model is correct: create a new index with the right dimension and re-ingest everything.
2. If the index is correct: pin the embedding model back to the one the index was built for.
3. Never truncate or pad vectors to fit.

Verify the fix

Upsert one probe vector and read it back with fetch. Then re-run the full ingest and confirm zero dimension errors in the logs.

Pinecone's index guides fix dimension at creation time against the embedding model; the troubleshooting pages cover upsert numeric handling. Diagnosis pattern follows the docs' support-evidence guidance. Verified against live docs 2026-09-26.