Open-source Deep Research
Open-source Deep Research for Perplexica Users
Bring broad cited discovery into agent workflows while keeping your research stack open-source.
01
Built for agent call sites
Perplexica-style workflows emphasize open-source answer finding. AutoSearch focuses on agent integration through MCP, making source discovery callable from coding assistants, research agents, and automation flows instead of only a separate search UI.
02
Source diversity for hard questions
Complex research rarely lives in one source type. AutoSearch can combine docs, repositories, papers, forums, video, and Chinese platforms, giving agents a wider evidence base before they produce conclusions or next-step plans.
03
LLM-decoupled by default
Teams can keep using their preferred host and model while AutoSearch manages retrieval across channels. That split is practical for experimentation, self-hosting, and workflows where answer generation and source discovery should evolve independently.
How it fits
AutoSearch belongs in the part of the stack where an agent needs source-backed research on demand. Perplexica can remain useful for open-source answer exploration, while AutoSearch gives MCP-capable hosts a direct way to collect cited evidence from 40 channels. The agent then decides how to analyze or report the material. This keeps the research path open, portable, and independent from any single LLM interface.
Try this prompt
Research open-source deep research tools for agent workflows.
Compare setup, channel coverage, and citation behavior with links.