Repolode
Submit
alphaXiv

alphaXiv/OpenResearch

A local-first harness and workspace that turns coding agents like Claude Code and Codex into research agents for literature review, experiments, and artifacts.

View on GitHubWebsite
Stars
7,333
Forks
492
Open issues
134
Last commit
6 hours ago

What is it?

OpenResearch is a desktop app and CLI (`orx`) for running research agents. It lets agents review literature, form hypotheses, run experiments, and produce research artifacts, working with coding agents such as Claude Code, Codex, OpenCode, and Cursor, as well as local models through LM Studio, Ollama, or a custom endpoint.

Agents can search papers via alphaXiv, bioRxiv, and PubMed, and run experiments on local machines or on remote compute including SSH, Slurm, Kubernetes, Modal, Ray, and Hugging Face Jobs. Every experiment is recorded in a local SQL database along with its logs, code, and artifacts, so agents can build on past results.

All projects, conversations, and experiment data are stored on the user's machine. Usage analytics are opt-out and exclude code, prompts, and file paths.

Key features

  • 01Works with Claude Code, Codex, OpenCode, Cursor, and Google Antigravity.
  • 02Connects to alphaXiv, bioRxiv, and PubMed for literature search.
  • 03Runs experiments locally or on SSH, Slurm, Kubernetes, Modal, Ray, and Hugging Face Jobs.
  • 04Stores experiment records, logs, and artifacts in a local SQL database.
  • 05Supports local models via LM Studio, oMLX, Ollama, or a custom endpoint.

Use cases

  • Having a coding agent survey recent papers before designing an experiment.
  • Running and tracking machine learning experiments across local and cloud compute.
  • Keeping research data and agent conversations on a local machine.
  • Letting agents reuse results from earlier experiments to decide next steps.

Topics

#ai-agents#research#llm#experiment-tracking#literature-review#local-first#claude-code#rust

FAQ

What is OpenResearch?
A local-first harness and workspace that turns coding agents like Claude Code and Codex into research agents for literature review, experiments, and artifacts. OpenResearch is a desktop app and CLI (`orx`) for running research agents. It lets agents review literature, form hypotheses, run experiments, and produce research artifacts, working with coding agents such as Claude Code, Codex, OpenCode, and Cursor, as well as local models through LM Studio, Ollama, or a custom endpoint.
Is OpenResearch free to use? What license does it have?
OpenResearch is open source under the MIT license. Check the license terms before using it in commercial products.
Is OpenResearch still maintained?
Actively maintained. The last commit to OpenResearch was on Oct 18, 2026.
What can I use OpenResearch for?
Having a coding agent survey recent papers before designing an experiment. Running and tracking machine learning experiments across local and cloud compute. Keeping research data and agent conversations on a local machine. Letting agents reuse results from earlier experiments to decide next steps.
Where can I get OpenResearch?
The source code and installation instructions are on GitHub: https://github.com/alphaXiv/OpenResearch

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects