Skip to navigation

Welcome to Rasen Developer API

Build, launch, and refine voice AI agents from your backend.

The Rasen Developer API gives a backend one workspace-scoped interface for the agent lifecycle: create and configure an agent, connect tools and phone routing, start calls, and read the evidence that each call produces.

Start with the path you need

GoalStart hereResult
Make the first requestQuickstartVerify a key and create an agent with a retry-safe request.
Start a campaignCampaignsCreate recipients, plan, admit and start outbound calls.
Build with an AI assistantBuild with AI assistantsAsk the docs a question, or connect Claude Code, Cursor or VS Code to the Docs MCP server.
Find an endpointAPI ReferenceInspect its request, response, errors, and code samples.
Prepare production codeIntegration checklistCheck credentials, retries, paging, revisions, and support data.

Build

CapabilityWhat you can do
Agent lifecycleCreate agents, edit drafts and variables, validate, publish, compare versions, and roll back.
VoicesList the voices a new agent starts with, or the voices an existing agent can select in its draft.
Tools and MCP serversAdd webhook tools, register MCP servers, discover tools, and attach them to a draft.
Tool credentialsStore write-only credentials, rotate versions, inspect references, and revoke a version.
KnowledgeAdd sources to the agent library and freeze ready or fallback content when publishing.
Hosted test-call linksEnable, disable and rotate a bounded test-call link.

Launch

CapabilityWhat you can do
Phone routingAssign a published agent to an existing workspace number for inbound or outbound calls.
CallsStart and end calls and use your own metadata to find them again.
CampaignsCreate, upload, plan, admit, start, pause, resume, retry and cancel campaigns.
Telephony connectionsManage carrier connections and import existing numbers.

Connect a supported carrier, import an existing number, then assign a published agent to it. Carrier number purchase and release stay with the provider.

Refine

CapabilityWhat you can do
Call resultsRead transcripts, recordings, extraction, summaries, and tool outcomes; export campaign results and transcripts by batch ID.
UsageRead usage by agent or call without deriving billing truth from raw events.
Change feedsPage through bounded call and usage changes and keep a local copy in sync.

The public API contains 104 operations. See API availability for each family and its requirements.

One stable public base

Every operation uses the configured dedicated API host under /v1:

https://api.v2.rasen.ai/v1

Requests use a workspace API key as a bearer token. The key cannot select another workspace, and there is no workspace-selector header. Continue with Authentication or make the first request in the Quickstart.