OpenSpeaker.ai
Agent data
ENVI
Start creating

Generation API · REST · asynchronous tasks

Build with OpenSpeaker.
One API for many AI workflows.

Integrate speech, dialogue, audio, Suno music and image generation through one base URL. The API uses the same premium-credit balance as the web application.

Open full in-app docsSee pricing

Task model

Create once. Receive a task ID. Poll or wait for a webhook.

Most generation APIs are asynchronous so long audio, music and image work can run without keeping the request open.

01

Create a task

POST the workflow payload. Include receive_url when webhook delivery is preferred.

02

Receive task_id

The create response returns the shared ID used to check status and retrieve output.

03

Poll or use a webhook

Call GET /v1/task/{task_id}, or let OpenSpeaker deliver the result to receive_url.

Generation endpoints

Source-verified production workflows.

This list intentionally excludes Kokoro, Kiki, legacy MiniMax Music, legacy Google Imagen and Video/Veo because they are not in the active public catalog.

ToolMethodPath
Text to SpeechPOST/v3/text-to-speech
Text to DialoguePOST/v3/text-to-speech/dialogue
Pronunciation DictionaryGET/v3/dictionaries
Pronunciation DictionaryPOST/v3/dictionaries
Pronunciation DictionaryPOST/v3/dictionaries/preview
Clone VoicePOST/v3/text-to-speech/voice-clone
AI Audio DubbingPOST/v1/task/dubbing
Voice ChangerPOST/v1/task/voice-changer
Voice IsolationPOST/v1/task/voice-isolate
Speech to TextPOST/v1/task/speech-to-text
Sound EffectsPOST/v1/task/sound-effect
AI Music GenerationPOST/v1s/task/music-generation
AI Image GenerationGET/v1i/models
AI Image GenerationPOST/v1i/task/price
AI Image GenerationPOST/v1i/task/generate-image

Safe example

Create speech with curl.

Replace $API_KEY and the voice ID with your account values. Never place a real key in browser source, a public repository or a prompt sent to a third party.

  • Text up to 1,000,000 characters per task
  • Speed from 0.5 to 1.5
  • Transcript, pronunciation dictionary and webhook options
curl
curl -X POST "https://api.openspeaker.ai/v3/text-to-speech" \
  -H "Content-Type: application/json" \
  -H "xi-api-key: $API_KEY" \
  -d '{
    "text": "A long-form narration starts here.",
    "voice_id": "provider:voice-id",
    "speed": 1,
    "receive_url": "https://example.com/webhooks/openspeaker"
  }'

Common endpoints

Tasks, credits, voices and health.

Use these endpoints to operate an integration: inspect history, check credits, find voices or delete tasks.

GET/v1/task/{task_id}

Poll one asynchronous task

GET/v1/tasks

List and filter task history

POST/v1/task/delete

Delete a task; refund where available

GET/v1/credits

Read the current credit balance

GET/v1/health-check

Check API availability

GET/v3/voices

Search the unified voice library

For AI agents and code generators

The API guide is machine-readable too.

Use the clean text guide or generation OpenAPI document to discover endpoints without crawling the authenticated interface.

/api-guide.txt/openapi-generation.json