wavespeed-ai/vibevoice is an advanced voice generation model for producing high-fidelity, natural, and expressive speech from text, with optional speaker/region-style control for more precise results and easy integration into real-world applications. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Idle
$0.015per run·~66 / $1
VibeVoice is a long-form text-to-speech (TTS) model designed to generate natural, podcast-like speech from transcripts, including multi-speaker conversations. It's built to stay coherent over long scripts while keeping each speaker's voice and speaking style consistent.
VibeVoice is most useful when you need dialogue, narration, or episode-length scripts rendered as speech. For background on the underlying model family, see the VibeVoice Technical Report and the Microsoft VibeVoice project page.
Long-form speech generation Handles extended transcripts (up to ~90 minutes in the long-form variant), useful for podcasts, audiobooks, and lecture-style narration.
Multi-speaker dialogue in one request Supports up to 4 speakers in a single generation, making it well-suited for interviews, panel discussions, and scripted conversations.
Consistent speaker identity across long scripts Designed to preserve each speaker's "voice" and conversational flow over long context windows.
Natural pacing and conversational delivery Optimized for dialogue-like speech (turn-taking, pauses, and rhythm) rather than robotic, sentence-by-sentence readouts.
Model-family support for low-latency streaming (variant-dependent) Some VibeVoice releases include a real-time streaming model optimized for fast first audio output; availability depends on the specific deployment/variant.
VibeVoice works best when your text looks like a real script:
Write it like a transcript, not a paragraph. Use short utterances, turn-taking, and punctuation that reflects how you want it spoken.
For multi-speaker dialogue, tag speakers clearly.
Common patterns include speaker tags like S1:, S2:, etc. If your wrapper expects a specific tag format (for example [S1] / [S2]), follow what the Playground examples show.
Keep overlap out of the script. If two speakers talk over each other in the transcript, the model may flatten it into a single line or produce unstable timing.
Use lightweight direction cues sparingly.
Short cues like (pause) or (laughs) may help with delivery, but results vary by model variant and deployment.
Example (single request, multi-speaker style):
S1: Welcome back. Today we're talking about shipping fast without breaking trust.
S2: The trick is to be explicit about trade-offs—especially in the UI.
S1: Let's start with a real example.
voice_id, pick one of the available built-in voices. For multi-speaker scripts, some deployments may apply fixed voices automatically; others may expose multiple voice selectors. Prefer what the Playground/UI schema provides.After you finish configuring the parameters, click Run, preview the result, and iterate if needed.
Minimum Pricing: $0.015 per run
Pricing is defined in this model's WaveSpeedAI configuration and is shown in the Playground cost preview before you run.
Best-effort language support varies by release. Many VibeVoice releases focus on English and Chinese; other languages may work inconsistently depending on the deployed speaker set.
Plan for long scripts. If you're generating a full episode, structure the transcript with clear segments (intro → sections → outro). If you hit instability, split into multiple runs and stitch audio in post.
Use responsibly. High-quality speech synthesis can be misused for impersonation or deceptive content. Only generate voices you have the rights and consent to use, and disclose AI-generated audio where appropriate.
Grab a WaveSpeedAI API key, then call POST https://api.wavespeed.ai/api/v3/wavespeed-ai/vibevoice with your input as JSON. The endpoint returns a prediction id. Start polling the result endpoint around every 2 seconds, increase the interval for long-running tasks, and stop on any terminal status. On completed, read output values from data.outputs. Examples for Vibevoice below.
set -euo pipefail
: "${WAVESPEED_API_KEY:?Set WAVESPEED_API_KEY}"
REQUEST_BODY=$(cat <<'JSON'
{
"text": "A clear example input",
"speaker": "Frank"
}
JSON
)
# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
-X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/vibevoice" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $WAVESPEED_API_KEY" \
-d "$REQUEST_BODY")
TASK=$(printf '%s' "$SUBMIT_RESPONSE" | jq 'if has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "$TASK" | jq -r '.id')
if [ -z "$PREDICTION_ID" ] || [ "$PREDICTION_ID" = "null" ]; then
printf 'Submission response did not contain a prediction id
' >&2
exit 1
fi
RESULT_URL=$(printf '%s' "$TASK" | jq -r '.urls.get // empty')
if [ -z "$RESULT_URL" ]; then
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/$PREDICTION_ID/result"
fi
# 2. Poll until the prediction finishes.
while true; do
RESPONSE=$(curl --silent --show-error --fail-with-body "$RESULT_URL" \
-H "Authorization: Bearer $WAVESPEED_API_KEY")
RESULT=$(printf '%s' "$RESPONSE" | jq 'if has("data") then .data else . end')
STATUS=$(printf '%s' "$RESULT" | jq -r '.status')
case "$STATUS" in
completed) printf '%s\n' "$RESULT" | jq '.outputs'; break ;;
failed|cancelled|timeout) printf '%s\n' "$RESULT" | jq . >&2; exit 1 ;;
created|processing) sleep 2 ;;
*) printf 'Unexpected status: %s
' "$STATUS" >&2; exit 1 ;;
esac
doneconst submitUrl = "https://api.wavespeed.ai/api/v3/wavespeed-ai/vibevoice";
const apiKey = process.env.WAVESPEED_API_KEY;
if (!apiKey) throw new Error('Set WAVESPEED_API_KEY');
async function requestJson(url, options = {}) {
const response = await fetch(url, options);
if (!response.ok) throw new Error(await response.text());
return response.json();
}
// 1. Submit the prediction.
const body = await requestJson(submitUrl, {
method: "POST",
headers: {
"Authorization": `Bearer ${apiKey}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"text": "A clear example input",
"speaker": "Frank"
}),
});
const task = body.data ?? body;
if (!task.id) throw new Error("Submission response did not contain a prediction id");
const resultUrl = task.urls?.get ||
`https://api.wavespeed.ai/api/v3/predictions/${task.id}/result`;
// 2. Poll until the prediction finishes.
while (true) {
const resultBody = await requestJson(resultUrl, {
headers: { "Authorization": `Bearer ${apiKey}` },
});
const result = resultBody.data ?? resultBody;
if (result.status === "completed") {
console.log(result.outputs);
break;
}
if (["failed", "cancelled", "timeout"].includes(result.status)) throw new Error(JSON.stringify(result));
if (!["created", "processing"].includes(result.status)) throw new Error("Unexpected status: " + result.status);
await new Promise(resolve => setTimeout(resolve, 2000));
}import json
import os
import time
from urllib.request import Request, urlopen
api_key = os.environ["WAVESPEED_API_KEY"]
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
payload = {
"text": "A clear example input",
"speaker": "Frank"
}
def request_json(url, data=None):
request = Request(url, data=data, headers=headers, method="POST" if data else "GET")
with urlopen(request) as response:
return json.load(response)
# 1. Submit the prediction.
body = request_json("https://api.wavespeed.ai/api/v3/wavespeed-ai/vibevoice", json.dumps(payload).encode())
task = body.get("data", body)
if not task.get("id"):
raise RuntimeError("Submission response did not contain a prediction id")
result_url = task.get("urls", {}).get("get") or f"https://api.wavespeed.ai/api/v3/predictions/{task['id']}/result"
# 2. Poll until the prediction finishes.
while True:
result_body = request_json(result_url)
result = result_body.get("data", result_body)
status = result.get("status")
if status == "completed":
print(result.get("outputs", []))
break
if status in {"failed", "cancelled", "timeout"}:
raise RuntimeError(result)
if status not in {"created", "processing"}:
raise RuntimeError(f"Unexpected status: {status}")
time.sleep(2)Vibevoice is a WaveSpeedAI model for audio generation, exposed as a REST API on WaveSpeedAI. wavespeed-ai/vibevoice is an advanced voice generation model for producing high-fidelity, natural, and expressive speech from text, with optional speaker/region-style control for more precise results and easy integration into real-world applications. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing. You can call it programmatically or try it from the playground above.
POST your input parameters to the model's REST endpoint (shown in the API tab of this playground) with your WaveSpeedAI API key in the Authorization header. Submission returns a prediction ID. Poll the result endpoint starting around every 2 seconds, increase the interval for long-running tasks, and stop on any terminal status. The playground generates production-oriented Python, JavaScript, and cURL examples with timeouts, transient-error handling, and safe GET retries. Full request/response shape is documented at https://wavespeed.ai/docs/docs-api/wavespeed-ai/vibevoice.
Vibevoice starts at $0.015 per run. That figure is the base price — the final charge scales with the parameters you set in the form (output size, length, count, references, or whatever knobs this model exposes), so a higher-quality or larger output costs more than a minimal one. The exact cost for your current input is shown live next to the Generate button before you submit, and the actual per-call charge is recorded on the prediction afterwards.
Key inputs: `speaker`, `text`. The full JSON schema (types, defaults, allowed values) is rendered above the Generate button and mirrored in the API reference at https://wavespeed.ai/docs/docs-api/wavespeed-ai/vibevoice.
Median end-to-end generation time on WaveSpeedAI is around 49 seconds per request, based on recent successful runs. Queue time varies with global demand; live status is visible in the prediction record.
Commercial usage rights depend on the model's license, set by its provider (WaveSpeedAI). The license summary appears on the model card above; see WaveSpeedAI's Terms of Service for platform-level conditions.