Can I use Laravel AI SDK for web search and embedding-backed features in Eloquent models?

Yes. The Laravel AI SDK includes embedding and web-search capabilities that you can use directly from Eloquent flows, with multi-provider support and structured outputs to make parsing results reliable. As Taylor Otwell explains, the SDK provides tools like web search and web fetch, plus a unified interface across providers and streaming/attachments for practical data workflows, so you can power data retrieval and analysis inside your Laravel app. Nuno Maduro emphasizes that embeddings and vector stores are built into the toolkit (including PGVector/Postgres integration), and that you can leverage cued agents and structured output to reason over documents or model results, all while keeping memory and conversation context manageable. In the release-focused videos, it’s highlighted that Eloquent integration includes embedding-related features that make vector-based queries feel native to your models, enabling operations like storing and querying embeddings alongside your Eloquent data. Laravel Daily notes that Laravel AI SDK becomes stable with Laravel 13 and points to semantic vector search (e.g., whereVector) as a supported feature, reinforcing that embedding-backed querying can be used in models without leaving the ORM layer. Overall, you can build embedding-backed features and web-search-enabled AI workflows inside Eloquent models by leveraging the SDK’s embeddings, PGVector support, structured outputs, and integrated tools, while benefiting from multi-provider orchestration and memory/agent capabilities.

• As Taylor Otwell explains, the SDK exposes built-in tools like web search and web fetch that augment LLMs with real‑world data retrieval within Laravel.
• Nuno Maduro points out that embeddings and vector stores are built into the SDK, with PGVector Postgres integration enabling efficient vector queries inside your app.
• The release-focused talks highlight that Eloquent integration includes embedding features, making vector-based queries feel like first‑class citizens in models.
• Laravel Daily notes that semantic vector search (including whereVector) is available in Laravel 13, underscoring native ORM-friendly embedding-backed querying.
• Across creators, the SDK’s multi-provider architecture and structured JSON outputs help you build robust, testable AI workflows that combine web data, embeddings, and model outputs in Laravel models.