Update speech
Update the agent's speech and call-audio knobs: voice speed, volume and emotion, carrier-side noise suppression for the caller's inbound audio on phone calls, and browser-mic noise suppression on web calls (test panel, demos, share page). Every field is optional; send only what changes. Use this when the user says things like 'speak slower' (ttsSpeed: 0.85), 'sound more cheerful' (ttsEmotion: 'happy'), 'quieter' (ttsVolume: 0.8), 'clean up background noise on phone calls' (noiseSuppression: one of the engines the field lists, or 'off' to disable), or 'clean up mic noise on web test calls' (noiseSuppressionWeb: true | false). noiseSuppressionStrength (0-100 whole number, default 70) tunes how hard the phone-side engine cleans: lower keeps more natural audio, higher removes more noise but can clip quiet speech and pump; inert when noiseSuppression is 'off' or 'Denoiser'. elTone ('natural' | 'expressive' | 'steady') sets the tone preset on the voices that have one, and applies to no other voice; the default voice engine uses ttsEmotion instead. sttLanguageHints lists the languages callers are expected to speak (codes from en, it, es, fr, de, pt, nl, ru, hi, ja) and only applies to the multilingual speech recognition model (set_stt_model): it biases which language each caller turn is heard as, so a short reply such as 'Va bene' is not heard as Dutch. The agent's voice pronounces each sentence it speaks in whichever listed language that sentence is written in, so list at least two languages for the voice to switch. The default is English; pass [] to return to it. Pass null for a field to unset it, back to the default.
Authorization
bearerAuth A personal access token from Settings > API tokens.
In: header
Header Parameters
A key you choose, such as a UUID, that makes this POST safe to retry. Within 24 hours the same key and request answers with the first successful answer and runs nothing, the same key with a different request answers 422 idempotency_key_reused, and a key whose request is still running answers 409 idempotency_key_in_use. A failed request keeps nothing, so the same key runs it again.
1 <= length <= 255Request Body
application/json
TypeScript Definitions
Use the request body type in TypeScript.
Response Body
application/json
application/problem+json
application/problem+json
application/problem+json
application/problem+json
application/problem+json
application/problem+json
application/problem+json
curl -X POST "https://example.com/operations/set_speech_settings" \ -H "Content-Type: application/json" \ -d '{ "agent": "string" }'{}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}{
"type": "http://example.com",
"title": "string",
"status": 0,
"detail": "string",
"code": "string",
"errors": [
{
"path": "string",
"message": "string"
}
],
"estimate": {},
"retryAfterSeconds": 0
}Set scenario environment POST
Set the runtime environment (audio-degradation) knobs on a test-harness scenario, how degraded the line the simulated caller drives the agent over is. Knobs: backgroundNoiseLevel (0-1) raises or lowers the caller ambience, and when the persona has no ambience of its own it plays office ambience, so asking for noise always produces noise; networkLatencyMs (0-2000ms) adds one-way delay; audioQuality (0.2-1) attenuates the caller's audio, 1 = clean. Out-of-range values are clamped, not rejected. VOICE scenarios only, a text scenario ignores all of them. Send only what changes; omitting a knob (or sending an empty environment) leaves the caller on a clean line, the default reproduces today's caller exactly. This configures the TEST HARNESS caller only, it never changes how the agent itself behaves.
Set STT model POST
Set which speech-to-text engine transcribes the caller on this agent. The STT model is versioned, the write lands on the current draft version (create a draft first if the current version is published). Pass null to clear the override and fall back to the platform default. Unlike the LLM model this affects transcription accuracy, language coverage, and keyterm boosting support, call get_agent (sections: ['model']) first if unsure what's currently set.