Bionic GPT
An open-source AI agent platform for on-premise or air-gapped deployment, with local models, private knowledge and internal API integrations.
Bionic GPT is a self-hosted AI agent platform for internal AI teams that need control over their models, company data and business systems. It runs on-premise, in a private cloud or in an air-gapped environment. Teams can build company-specific workflows on its existing workspace and agent runtime rather than assemble the underlying platform themselves.
Agents can retrieve information from private datasets, work with uploaded files, call approved tools and execute code in a sandbox. They can carry context through multi-step tasks across business systems and save generated files and other outputs. Reusable skills hold domain instructions and templates for work that needs a repeatable method.
Bionic accepts OpenAPI specifications to connect services such as Gmail, Salesforce and internal applications without a custom agent for each integration. Its chat workspace includes conversation history, while team permissions, identity controls and audit trails support shared use inside an organisation.
The platform works with local models, private inference endpoints and approved hosted models. A deployment using local models and internal systems can stay inside the organisation's perimeter; hosted models and external integrations send requests outside it. Bionic also offers a fully managed cloud deployment.
The open-source foundation uses the Apache 2.0 license and is written in Rust. Docker Compose supports local evaluation and small pilots, while its production infrastructure centres on Kubernetes.
