Everything between query and answer.

One API to search the web, fetch web pages and documents as clean markdown, ask AI questions about a page, or run deep multi-source research.

Get Started See Pricing

Four tools, one API.

REST or MCP, one key, clean markdown out.

Search the web and get page content, not just snippets. Choose a search depth to balance speed and detail.

POST /v1/search 1–2 credits

Fetch URLs and get clean markdown. Works with web pages, PDFs, DOCX, and other document formats.

POST /v1/fetch 2 credits per URL

Fetch a page and ask AI about it. Pricing, contact info, specs - describe what you need and get a written answer.

POST /v1/extract 10 credits

Deep research on any topic. Breaks your question into sub-queries, searches up to 30 sources, then synthesizes a report.

POST /v1/research 50 credits

Tap a client for a 30-second setup guide. Any MCP client that supports remote (HTTP) servers works.

Replace ay_live_... with your key from the dashboard. It travels as an HTTP header, so the model never sees it.

Projects and platforms shipping Sofya as a native integration.

Listed MCP server: all four tools for Copilot and VS Code agents.

Search provider in the browser extension for chatting with local AI models.

Built-in search tool for agents in the open-source RAG engine.

Marketplace plugin: all four tools for Dify agents and workflows.

Web search and page fetch provider for ByteDance's deep research agent.

Search + fetch backend for the pi coding agent.

Built-in web-search backend for the terminal coding agent.

Official Sofya piece: all four tools as no-code automation steps.

Web-search provider and a native sofya-research model for AI chat.

SDK to build, run, and manage agent platforms.

Built-in search engine for AI-powered deep research reports.

Search provider for the web-search MCP server used by coding agents.