Skip to main content
The Vida API exposes the same core resources used to operate the AI Agent OS. Use it to configure conversational and Computer Agents, connect communications, create and track work, inspect results, and integrate Vida with your systems.

​
What you can build

  • Agent configuration: Discover available models, voices, functions, and applications; edit staging; save versions; run experiments; and publish verified changes.
  • Communications: Place and receive calls, send and receive messages, manage phone and SIP connectivity, and inspect complete conversation history.
  • Tasks and workflows: Queue communication work, define goals and retry cadences, assign Computer work, and manage repeating or one-off schedules.
  • Computer Agents: Provision and support Agents that use skills, managed credentials, channels, browsers, files, memory, and reusable helpers.
  • Contacts and objectives: Maintain customer context and ongoing work without inventing a separate system of record.
  • Reporting and operations: Query logs, conversations, usage, Task outcomes, and time-series metrics.
  • Platform administration: Onboard organizations and Agents, manage optional reseller or partner hierarchies, apply templates, control access, configure inbound email and webhooks, and manage white-label domains when those capabilities are enabled.

​
How the reference is organized

The generated endpoint pages define exact request and response schemas. The API guides explain multi-step workflows, resource relationships, and verification. Start with Getting started, then use the guide for the resource you are operating. Treat accepted writes, queued Tasks, and started lifecycle jobs as pending until a follow-up read proves the intended state or outcome.

​
AI-agent integrations

If an AI agent will build or operate your integration, give it the Vida API Skill. The OpenAPI document remains authoritative for endpoint shape; the skill supplies account-scoping, sequencing, safety, and verification guidance. The current machine-readable definition is available from the OpenAPI Definition link in the navigation.