AIPOCH Open-Science
An AI research workbench runs on macOS, Windows and Linux, combining Python and R analysis, compatible local models and versioned research outputs.

AIPOCH Open-Science is a desktop AI research workbench for scientists who want to keep investigations, analysis and supporting evidence in one project. Its agents work with research files and executable tools to produce outputs you can inspect alongside the records behind them.
- Research agents: A coordinator can delegate focused tasks to specialist agents for literature review, data analysis and other research work, subject to your workspace permissions.
- Python and R analysis: Persistent sessions retain variables between steps, while shell tools and supported remote SSH compute handle additional research tasks.
- Scientific resources: Built-in connectors cover literature, biomedical databases, genomics, chemistry and clinical research; reusable skills and compatible custom MCP connectors extend the available tools.
- Research records: Reports, tables, figures and notebooks have version histories linked to available inputs, code, execution records and environment details. Optional Reviewer checks flag unsupported claims, inconsistencies and missing evidence, but don't certify scientific correctness.
The app is free and open source under Apache 2.0, and runs on macOS, Windows and Linux. It supports cloud providers such as OpenAI and Anthropic, plus compatible gateways with presets for Ollama, LM Studio, llama.cpp and vLLM. Model compatibility depends on the selected agent backend and API protocol.
Project state, uploads, sessions and generated artifacts stay on your computer by default. Calls to external model providers, scientific connectors, web search or remote compute can send data off the device, and external services may charge separately. You can export completed sessions as portable .science packages for review, handoff and archiving.