Skip to content

VibeVoice 7B

Integrate

Model id vibevoice/7b
# Run the model; wait up to 20 s for the finished result in the same call
curl -X POST "https://api.doitong.com/v1/run/vibevoice/7b?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "script": "A red fox trots through fresh snow at sunrise, soft golden light, shallow depth of field",
    "speakers": [
      {
        "preset": "Frank [EN]"
      },
      {
        "preset": "Carter [EN]"
      }
    ]
  }'

# Still "queued" or "running"? Poll with the id from the response:
curl "https://api.doitong.com/v1/requests/REQUEST_ID?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY"

This request: ≈$0.04 40 API credits Run it on this page

Get API key
Microsoft Text to Speech #text-to-speech#multi-speaker#podcast

Larger VibeVoice model for multi-speaker dialogue, with scripts up to 30000 characters

Microsoft · Audio

Your first result goes here

Build a request below and run it with your API key. Outputs are copied to our storage, kept forever or 7 days for free.

Build a request

Controls are generated from the model schema. The code and the price update as you change them.

•required 88 / 30000

Speech input script, with optional 'Speaker X:' labels to distinguish speakers in a dialogue.

•required

Speakers are inferred automatically using the script or supplied voice samples; provide a list to specify them explicitly.

Advanced cfg_scale · seed · enable_safety_checker
default 1.3

CFG (Classifier-Free Guidance) strength determines how closely generation follows the text, with stronger adherence at higher settings.

Random seed that makes generation reproducible.

POST /run/vibevoice/7b?wait=20 quote → $0.04
# Run the model; wait up to 20 s for the finished result in the same call
curl -X POST "https://api.doitong.com/v1/run/vibevoice/7b?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "script": "A red fox trots through fresh snow at sunrise, soft golden light, shallow depth of field",
    "speakers": [
      {
        "preset": "Frank [EN]"
      },
      {
        "preset": "Carter [EN]"
      }
    ]
  }'

# Still "queued" or "running"? Poll with the id from the response:
curl "https://api.doitong.com/v1/requests/REQUEST_ID?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY"

Run this request

Your key stays in this browser: it is sent only to the Doitong API and saved (localStorage) only if you tick Remember. Get an API key

About VibeVoice 7B

Turns a `script` of up to 30000 characters into speech, with 'Speaker X:' prefixes for dialogue and a required `speakers` list for the voices. `cfg_scale` ranges from 1 to 2 and `seed` allows reproducible output. It is the 7B version of the family, with a shorter text limit than the 1.5B model.

Parameters

NameType DefaultAllowed Description
scriptrequired
Typestring
≤ 30,000 characters
Speech input script, with optional 'Speaker X:' labels to distinguish speakers in a dialogue.
speakersrequired
Typeobject[]
Speakers are inferred automatically using the script or supplied voice samples; provide a list to specify them explicitly.
speakers[].presetoptional
Typeenum
DefaultAlice [EN]
Alice [EN]Carter [EN]Frank [EN]Mary [EN] (Background Music)Maya [EN]Anchen [ZH] (Background Music)Bowen [ZH]Xinran [ZH]
Fallback voice preset for a speaker without `audio_url`; supplying that audio sample takes precedence over the preset.
speakers[].audio_urloptional
Typeurl · audio
Voice sample audio file URL; supplying it overrides `preset`.
seedoptional
Typeinteger | null
Random seed that makes generation reproducible.
cfg_scaleoptional
Typenumber
Default1.3
1 – 2
CFG (Classifier-Free Guidance) strength determines how closely generation follows the text, with stronger adherence at higher settings.
enable_safety_checkeroptional
Typeboolean
Defaulttrue
Screening can be relaxed by setting false, provided the caller is authorized.
required

Output

GET /requests/req_8k2m4q7x1z9c3v5b6n0p2r4t
{
  "id": "req_8k2m4q7x1z9c3v5b6n0p2r4t",
  "object": "request",
  "model": "vibevoice/7b",
  "status": "succeeded",
  "created_at": "2026-10-20T12:00:00Z",
  "started_at": "2026-10-20T12:00:02Z",
  "completed_at": "2026-10-20T12:00:09Z",
  "queue_position": null,
  "cost": {
    "credits": 40,
    "usd": "0.04",
    "charged": 40,
    "refunded": 0
  },
  "input": {
    "script": "A red fox trots through fresh snow at sunrise, soft golden light, shallow depth of field",
    "speakers": [
      {
        "preset": "Frank [EN]"
      },
      {
        "preset": "Carter [EN]"
      }
    ]
  },
  "output": {
    "audio": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.mp3",
      "content_type": "audio/mpeg",
      "file_name": "0-output.mp3",
      "file_size": 1843200
    },
    "duration": 5,
    "sample_rate": 24000,
    "generation_time": 5.6,
    "rtf": 0.53
  },
  "outputs": [
    {
      "type": "audio",
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.mp3",
      "content_type": "audio/mpeg",
      "file_size": 1843200,
      "duration": 12,
      "stored": true
    }
  ],
  "error": null,
  "metrics": {
    "inference_seconds": 6.8
  },
  "urls": {
    "self": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t",
    "cancel": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t/cancel"
  }
}
Your files, your choiceEvery outputs[].url is a copy on our storage: kept forever (paid per GB-month) or 7 days for free, by your account setting or ?storage=.
Same shape on every modeltype, url, content_type, width, height and duration, for images, video and audio alike. output is the model’s own result object.
Fields of output
audio
object {url, content_type, file_name, file_size}Audio file containing the synthesized speech.
duration
numberLength of the audio output, measured in seconds.
sample_rate
integerSampling rate of the audio output.
generation_time
numberAudio generation time, measured in seconds.
rtf
numberReal-time factor calculated as generation_time divided by audio_duration; smaller values indicate better performance.

More models: Text to Speech

For AI agents: this page as Markdown is at https://doitong.com/en-US/developers/models/vibevoice/7b.md. All models and guides: https://doitong.com/developers/llms.txt