OpenResearch is a local-first workspace that transforms coding agents like Claude Code, Codex, OpenCode, and Cursor into autonomous research agents.
It uses a git-native experiment tree and isolated worktrees to manage parallel research directions, ensuring that experiments are reproducible and evidence is traceable.
- 01
Isolated git worktrees for parallel agent sessions
- 02
Git-native experiment tree for lineage tracking
- 03
Support for Claude Code, Codex, OpenCode, and Cursor
- 04
Flexible compute options including local, SSH, cluster, and managed cloud
- 05
Local model integration via OpenCode (e.g., Ollama, LM Studio)
Free
The software is open-source (MIT License) and free to use locally. Costs are only incurred if you choose to use paid model providers or rent remote compute, including alphaXiv's optional managed compute service.
Best for
Researchers and developers who run iterative ML experiments, need to manage multiple research directions simultaneously, and require strict reproducibility for their code and results.
Not ideal for
Users whose research does not involve code repositories, such as pure literature reviews or qualitative theory papers, or those seeking a simple chatbot for summarizing papers.
A powerful, rigorous tool for serious research automation that solves the 'bookkeeping' problem of AI experimentation by leveraging git primitives, making it an essential utility for those who need to track the provenance of their research claims.
Version 0.2.0 was released on September 11, 2026, introducing Windows support, a reworked skill picker, and fixes for OpenCode configuration truncation.
As of
Is OpenResearch free?+
Yes - OpenResearch is Free. The software is open-source (MIT License) and free to use locally. Costs are only incurred if you choose to use paid model providers or rent remote compute, including alphaXiv's optional managed compute service.
Is OpenResearch open source?+
OpenResearch is not listed as open source on ai.dosa.dev - it is Free.
Who is OpenResearch best for?+
Researchers and developers who run iterative ML experiments, need to manage multiple research directions simultaneously, and require strict reproducibility for their code and results.
Who is OpenResearch not ideal for?+
Users whose research does not involve code repositories, such as pure literature reviews or qualitative theory papers, or those seeking a simple chatbot for summarizing papers.
What are the best OpenResearch alternatives?+
The closest OpenResearch alternatives on ai.dosa.dev are Jules, OpenHands, SWE-agent - all listed under Autonomous & Async Agents.
Who makes OpenResearch?+
OpenResearch is developed by alphaXiv. It is listed in the Autonomous & Async Agents category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.