AI runtime infrastructure
Documentation
Learn how Zyntry runtimes connect applications to models, providers, knowledge, tools, integrations, security controls, and telemetry.
Quickstart
Describe the AI feature you are building, review the runtime plan, and create a governed execution boundary.
Invoke a runtime
Send a server-side request with a scoped Bearer API key and runtime_id. Runtime policy decides the eligible models, sources, tools, and actions.
Knowledge and tools
Attach private documents, approved integrations, or MCP capabilities only when the workflow needs them.