Embedding model

Updated May 23, 2026

Dear Diary uses FastEmbed for local embedding generation. The default model is sentence-transformers/all-MiniLM-L6-v2, which provides a good balance of quality and performance.

Model download

On first use, the embedding model is downloaded automatically to a local cache. This may take a few moments depending on network speed.

Alternative models

Configure a different model via the EMBEDDING_MODEL environment variable. Consult the FastEmbed documentation for available models.