doppar / embeds
Semantic search for Doppar models: attribute-driven, locally-computed embeddings and similarity queries, no external API required.
Requires
- php: ^8.5
- doppar/ai: 4.*
Requires (Dev)
- doppar/framework: 4.*
- mockery/mockery: ^1.6
- phpstan/phpstan: ^2.1
- phpunit/phpunit: ^13.3
- predis/predis: ^3.3
Suggests
- predis/predis: Required to use the "redis" embeds driver.
Provides
None
Conflicts
None
Replaces
None
This package is auto-updated.
Last update: 2026-09-12 17:55:53 UTC
README
Doppar Embeds
Doppar Embeds adds semantic search to your models with a single attribute. Mark a column with #[Embeds], and Doppar keeps a numeric representation of its meaning — a vector — in sync automatically,
so Product::whereSimilarTo('description', 'a durable waterproof backpack') finds "rugged daypack for hiking in the rain" even though the two share almost no words. Everything runs locally: embeddings are computed on your own server using a small local model, with no external API call, no API key, and no per-request cost.
Documentation
Read the documentation from doppar official site Doppar Embeds
Contributing
Thank you for considering contributing to the Doppar framework! The contribution guide can be found in the Doppar documentation.
Code of Conduct
In order to ensure that the Doppar community is welcoming to all, please review and abide by the Code of Conduct.
Security Vulnerabilities
Please review our security policy on how to report security vulnerabilities.
License
The Doppar framework is open-sourced software licensed under the MIT license.