Astra DB: collection vector dimensions must match your embedder or inserts fail

Astra DB: collection vector dimensions must match your embedder or inserts fail

TL;DR

When you create an Astra DB collection with a vector dimension, that number has to match the output dimension of the embedding model you actually use. A mismatch (for example a 1536-dimension collection with a 768-dimension embedder) fails at insert time, and the error reads like a data problem rather than a configuration one. The Genkit Astra DB plugin docs call it out: "Ensure that you have set up your collection with the correct number of dimensions for the embedder that you are using." Know your model's output size (text-embedding-ada-002 is 1536, NV-Embed-QA is 1024) before creating the collection, because the dimension is fixed at collection creation.

Use this when you run into the situation in the title.

When not to use this skill: unrelated tasks. It covers only the procedure above.

Compatibility

The steps above apply to the commands named in them. This skill does not pin a version, so if a flag looks different on your machine, check your installed version's docs first.

Details

If you use Astra DB Vectorize instead, you skip the embedder entirely: the collection's vectorize integration generates the vectors, and attaching your own embedding model is wrong.

When you create an Astra DB collection with a vector dimension, that number has to match the output dimension of the embedding model you actually use. A mismatch (for example a 1536-dimension collection with a 768-dimension embedder) fails at insert time, and the error reads like a data problem rather than a configuration one. The Genkit Astra DB plugin docs call it out: "Ensure that you have set up your collection with the correct number of dimensions for the embedder that you are using." Know your model's output size (text-embedding-ada-002 is 1536, NV-Embed-QA is 1024) before creating the collection, because the dimension is fixed at collection creation. If you use Astra DB Vectorize instead, you skip the embedder entirely: the collection's vectorize integration generates the vectors, and attaching your own embedding model is wrong. Verified answer: When you create an Astra DB collection with a vector dimension, that number has to match the output dimension of the embedding model you actually use. A mismatch (for example a 1536-dimension collection with a 768-dimension embedder) fails at insert time, and the error reads like a data problem rather than a configuration one. The Genkit Astra DB plugin docs call it out: "Ensure that you have set up your collection with the correct number of dimensions for the embedder that you are using." Know your model's output size (text-embedding-ada-002 is 1536, NV-Embed-QA is 1024) before creating the collection, because the dimension is fixed at collection creation. If you use Astra DB Vectorize instead, you skip the embedder entirely: the collection's vectorize integration generates the vectors, and attaching your own embedding model is wrong. Source: https://github.com/genkit-ai/docsite/blob/HEAD/src/content/docs/docs/integrations/astra-db.mdx