Skip to main content
POST
/
generate
/
audio
Generate audio
curl --request POST \
  --url https://api.horizon.new/v1/generate/audio \
  --header 'Content-Type: application/json' \
  --data '{
  "prompt": "<string>",
  "voice": "<string>",
  "tempo": 135,
  "durationSeconds": 60,
  "webhookUrl": "<string>"
}'
{
  "jobId": "job_01hx9q9",
  "status": "queued",
  "statusUrl": "<string>",
  "result": {},
  "etaSeconds": 123,
  "jobType": "extract/pdf"
}
Render high-quality speech or music stems based on a descriptive prompt. Audio is delivered in both lossless (.wav) and compressed (.mp3) formats for immediate use in your products or marketing materials.

​
Request body

  • prompt — Required narrative or sonic description.
  • voice — Optional voice preset slug for speech synthesis.
  • tempo — Optional beats per minute when generating music; ignored for spoken word.
  • durationSeconds — Optional cap on clip length (up to 300); defaults to 60.
  • webhookUrl — Optional HTTPS URL Horizon should call when the audio job finishes.

​
Sample request

curl https://api.horizon.new/v1/generate/audio \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Warm, confident narrator welcoming new Horizon users.",
    "voice": "serene-female",
    "durationSeconds": 30
  }'

​
Response

Returns 202 Accepted with a jobId, statusUrl, and previews for each format. Completed jobs include waveform analysis (RMS, LUFS) so you can loudness-match clips programmatically.

​
Notes

  • When you omit voice, the generator auto-selects a voice that matches the prompt’s tone.
  • To get subtitles alongside speech, add "includeCaptions": true to the request body.
  • Poll GET /jobs/{jobId} (same destination as statusUrl) to check progress or retrieve the audio bundle later.

​
x402 flow

Audio rendering is likewise defended by Coinbase’s x402 protocol. When additional payment is required, Horizon returns a structured 402 response: 
HTTP/1.1 402 Payment Required
Content-Type: application/json

{
  "x402Version": 1,
  "accepts": [
    {
      "scheme": "exact",
      "network": "base-sepolia",
      "maxAmountRequired": "150000",               // $0.15 in 6‑decimal USDC
      "resource": "POST /generate/audio",
      "description": "Horizon audio generation",
      "mimeType": "application/json",
      "payTo": "0xYourReceivingWallet",
      "maxTimeoutSeconds": 180,
      "asset": "0xYourUSDCContract",
      "extra": {
        "name": "USDC",
        "version": "1"
      }
    }
  ],
  "error": null
}
To unlock the job:
  1. Send the challenge to your facilitator (see Client & server responsibilities) and call /verify and /settle.
  2. Replay the request with the facilitator’s Base64 payload inside the standard X-PAYMENT header: 
    curl https://api.horizon.new/v1/generate/audio \
  -H "Content-Type: application/json" \
  -H "X-PAYMENT: eyJ4NDAyVmVyc2lvbiI6MSwic2NoZW1lIjoiZXhhY3QiLC4uLn0=" \
  -d '{ ...original body... }'
  3. Horizon validates the proof (locally or via facilitator). Successful responses may include X-PAYMENT-RESPONSE with the settlement record.
See the Quickstart for sellers for facilitator setup guidance before you roll out paid audio generation.

Body

application/json
​
prompt
string
required
​
voice
string
​
tempo
integer
Required range: 30 <= x <= 240
​
durationSeconds
integer
default:60
Required range: 1 <= x <= 300
​
webhookUrl
string<uri>

Webhook to call when the audio job completes.

Response

Audio generation job accepted

​
jobId
string
required
Example:

"job_01hx9q9"

​
status
enum<string>
required
Available options:
queued,
processing,
completed,
failed
​
statusUrl
string<uri>
required

Canonical link to GET /jobs/{jobId} for this job.

​
jobType
string
required
Example:

"extract/pdf"

​
result
object | null

Present when the job completes synchronously.

​
etaSeconds
integer | null

Estimated seconds until completion.