Skip to content

deep-research-client logo

deep-research-client

A CLI tool for AI-powered deep research across multiple providers.

Quick Start

# Install
pip install deep-research-client

# Set up an API key (e.g., Perplexity)
export PERPLEXITY_API_KEY="your-key"

# Run a research query
deep-research-client research "Research the human CFAP300 gene including molecular function, disease associations, and evolutionary conservation"

Or run without installation:

uvx deep-research-client research "Analyze recent advances in solid-state battery technology for electric vehicles, focusing on materials science breakthroughs since 2023"

Features

  • Multiple Providers: OpenAI, Perplexity, Edison, Consensus, Cyberian
  • Smart Caching: Avoid expensive re-queries
  • Templates: Create reusable research queries with variables
  • Model Selection: Choose speed vs comprehensiveness
  • Citations: All results include source references

Supported Providers

Provider Best For Speed
OpenAI Comprehensive reports Slow
Perplexity General web research Fast-Slow
Edison Scientific literature Slow
Consensus Academic papers Fast
Cyberian Agent-based deep research Very Slow

Documentation

This documentation follows the Diátaxis framework:

Tutorials (Learning)

Step-by-step guides for getting started:

How-to Guides (Tasks)

Solve specific problems:

Reference (Information)

Technical details:

Explanation (Understanding)

Background and concepts:

Common Commands

# Research with default provider
deep-research-client research "Review the current evidence on microplastics in human tissue and potential health implications"

# Use specific provider for scientific literature
deep-research-client research "Summarize mechanisms of ferroptosis in cancer therapy" --provider edison

# Save comprehensive report to file
deep-research-client research "Analyze the state of nuclear fusion research including ITER progress and private sector developments" --output fusion-report.md

# List available providers
deep-research-client providers

# List available models
deep-research-client models

# Check cache
deep-research-client list-cache