https://voice-api.relationsai.com/v1 with a rai_test_ key, which never places a real call. It works once the API is enabled for your workspace (private preview: support@relationsai.com). Your key stays in this tab's memory; live keys are refused.Agents
AI voice agents. Agents are created, configured and published in the console; the API reads them and tells you whether one is ready to place calls. Agent editing through the API is planned.
List agents
Lists the workspace's agents, newest first. Agents are created, configured and published in the console; the API reads them. Deleted agents are not listed.
published is checked live for each page, so it can be null when that check is not available at
that moment. Use the readiness endpoint before placing calls with an agent.
GET /v1/agents · Operation ID: listAgents
query Parameters
limitMaximum number of objects to return, from 1 to 100. Defaults to 25.
cursorThe next_cursor of the previous page. Send it with exactly the same filters as the request that returned it; a changed filter set is 400 invalid_cursor.
List agents › Responses
A page of agents.
objectAlways list.
has_moreTrue when more objects exist after this page.
next_cursorPass as cursor (with the same filters) to fetch the next page. Null on the last page.
Agents on this page.
Retrieve an agent
Returns one agent. An agent that was deleted, belongs to another workspace, or is the test-mode Sample agent read with a live key is 404 resource_not_found.
GET /v1/agents/{agent_id} · Operation ID: getAgent
path Parameters
agent_id^agent_[A-Za-z0-9_-]… · requiredThe agent id (agent_...).
Retrieve an agent › Responses
The agent.
id^agent_[A-Za-z0-9_-]… · requiredAgent id (agent_...).
objectAlways agent.
livemodeMode of the key that read it. Agents are shared configuration, readable in both modes.
nameAgent name.
descriptionWhat the agent is for, as written in the console. Null when not set.
statusWhether the agent is switched on in the console. active agents can place calls; paused and draft agents cannot. Open enum: unknown values mean the agent cannot be assumed callable.
publishedTrue when the agent has a published version that calls use. Null when the publication state could not be read right now; use the readiness endpoint for a definite answer.
published_versionNumber of the published version, when known.
is_defaultTrue for the workspace's default agent.
sampleTrue only for the built-in Sample agent that exists in test mode.
created_atWhen the agent was created.
updated_atWhen the agent's configuration last changed, when known.
Check whether an agent can place calls
Checks, right now, everything about the agent and the workspace that would stop a call being placed, and returns every problem found (not just the first). An agent is ready when it is switched on, fully set up and published, has an outbound phone number, and the workspace has calling set up, an active account, spend headroom and enough balance.
Readiness does not check per-destination rules (the do-not-call list, consent, calling hours and the per-number daily limit); those are applied to each call. A ready agent can therefore still be refused for a particular number, and readiness can change at any time.
GET /v1/agents/{agent_id}/readiness · Operation ID: getAgentReadiness
path Parameters
agent_id^agent_[A-Za-z0-9_-]… · requiredThe agent id (agent_...).
Check whether an agent can place calls › Responses
The readiness result.
objectAlways agent_readiness.
livemodeMode of the key that asked.
agent_idThe agent checked.
readyTrue when reasons is empty.
Every reason the agent cannot place calls. Empty when ready.
checked_atWhen the check ran.