AI-powered deep research agent. Open-source alternative to OpenAI Deep Research.
Project description
DeepWorm 🐁
AI deep research tool that searches the web, reads sources, and synthesizes findings into comprehensive reports.
Quick Start
pip install deepworm
Set up an API key (choose one)
# Option 1: OpenAI
export OPENAI_API_KEY="sk-..."
# Option 2: Google Gemini
export GOOGLE_API_KEY="AIza..."
# Option 3: Anthropic Claude
export ANTHROPIC_API_KEY="sk-ant..."
Run a research
deepworm "what are the latest advances in quantum computing?"
That's it. DeepWorm will:
- Search the web for relevant sources
- Read and analyze each source
- Generate followup questions
- Go deeper recursively
- Produce a final report with citations
Command Line Options
# Control research depth
deepworm "topic" --depth 3 --breadth 5
# Save report to file
deepworm "topic" --output report.md
# Choose provider
deepworm "topic" --provider google
# Compare topics
deepworm compare "Python" "Rust"
# Interactive mode
deepworm interactive
# See all options
deepworm --help
Python API
from deepworm import Researcher
researcher = Researcher(provider="openai")
result = researcher.research("your topic")
print(result.report)
Supported Providers
| Provider | Env Variable | Models |
|---|---|---|
| OpenAI | OPENAI_API_KEY |
gpt-4o, gpt-4o-mini |
GOOGLE_API_KEY |
gemini-pro, gemini-flash | |
| Anthropic | ANTHROPIC_API_KEY |
claude-3.5, claude-3 haiku |
| Ollama | (none, local) | llama3, mistral |
Configuration
Create deepworm.yaml in your project root:
provider: google
model: gemini-pro
depth: 3
breadth: 5
output_dir: ./reports
Key Features
- Multi-provider LLM support (OpenAI, Google, Anthropic, Ollama)
- Recursive deep research with configurable depth and breadth
- Markdown reports with citations and source links
- Comparison mode for side-by-side analysis
- Research chains for multi-step investigations
- Built-in retry and rate limiting
- DuckDuckGo search (no API key needed for search)
- Interactive mode for exploratory research
- Research history and caching
- Template system for repeatable research
- Rich terminal output with progress tracking
- Extensible plugin system
License
MIT
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
deepworm-1.5.0.tar.gz
(317.5 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
deepworm-1.5.0-py3-none-any.whl
(243.7 kB
view details)
File details
Details for the file deepworm-1.5.0.tar.gz.
File metadata
- Download URL: deepworm-1.5.0.tar.gz
- Upload date:
- Size: 317.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.9.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
85b753628eb102abae5e1b97b89b6b42e840549eb14dc66c3bc0baaccfba22bc
|
|
| MD5 |
8127a8818f9634ea999f1734c2d4cd65
|
|
| BLAKE2b-256 |
2100584550bf073663568367e7e3b4b8bd3157f92c59fc0a4bc4bb6bf977e735
|
File details
Details for the file deepworm-1.5.0-py3-none-any.whl.
File metadata
- Download URL: deepworm-1.5.0-py3-none-any.whl
- Upload date:
- Size: 243.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.9.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ea7217bae383e568f82755c30e5f6dc878fee15d8d436c99301a2edf18abf38b
|
|
| MD5 |
2764051d9a2023bd7eec199b2f12af4c
|
|
| BLAKE2b-256 |
1341855760e68fdafee13730d25cf2f50cefe17d411c24d899859f425697a3e2
|