Developers query indexed content programmatically and wire retrieval into custom chat UIs, support bots, and agent tool calls — without building crawl, chunk, and embedding pipelines from scratch.
Generic LLMs guess. Agents that search first reply with context from your pricing pages, policies, and product copy. Chat With Your Site indexes the same content you show visitors and exposes it as a search endpoint your code can call.
Agent tool call: search_site("refund policy")
Your support bot calls the search API, receives the relevant policy chunks, and drafts a reply grounded in your live terms — not a generic LLM guess.
Read the API reference for request formats, authentication, and example calls.
-
Drop-in retrieval
Skip maintaining scrapers, vector stores, and sync jobs — use one pipeline from live site to searchable chunks.
-
Branded chat in your app
Build your own UI while we handle indexing behind the scenes.
-
Copilots and onboarding flows
Embed search in SaaS so new users ask setup questions and your agent retrieves answers from public docs and changelog.
- Is this a replacement for building RAG myself?
- For many teams, yes. We handle crawling, chunking, embeddings, and retrieval so your app calls search and focuses on the agent experience.
- How do agents call search?
- Use the search API from your backend or agent framework before generating a reply, passing keywords from the user question.
- What happens when my site changes?
- Re-crawls refresh the index so every integration searches current content instead of a stale export.
Crawl, index, and chat with your content in minutes — free to start.