Breaks
Insert explicit pauses into KugelAudio speech with SSML-style break tags: syntax, accepted duration values, how durations are snapped, and their limits.
On this page
Use <break> tags to insert an explicit silence at a specific point:
before a verification code, between list items, or wherever punctuation alone
doesn't give you the pause you want.
Your total is <break time="400ms"/> forty-two euros.Syntax
| Form | Meaning |
|---|---|
<break time="300ms"/> | Explicit duration in milliseconds |
<break time="1.5s"/> | Explicit duration in seconds (decimals allowed) |
<break strength="medium"/> | SSML strength preset (none, x-weak, weak, medium, strong, x-strong) |
<break/> | Default pause (200 ms) |
Use the self-closing form with at most one attribute. Attribute values may
use single or double quotes. A tag with both time and strength is not
recognized and stays in the text, and an unknown strength value becomes the
default 200 ms pause.
Durations are snapped
The model is trained on three discrete pause lengths: 200 ms, 400 ms, and 500 ms. Whatever you request is snapped to the nearest trained value (ties resolve to the longer pause), so the effective pauses are:
| You write | You get |
|---|---|
time="0ms" | no pause |
More than 0ms, less than 300ms | 200 ms |
300ms to less than 450ms | 400 ms (300 ms ties → 400 ms) |
450ms and up | 500 ms (450 ms ties → 500 ms) |
strength="x-weak" / "weak" / "medium" | 200 ms |
strength="strong" / "x-strong" | 500 ms |
strength="none" | no pause inserted (not snapped up) |
Need a silence longer than 500 ms? Chain tags:
<break time="500ms"/><break time="500ms"/> ≈ 1 s.
Example
audio = client.tts.generate(
text='Welcome to KugelAudio. <break time="400ms"/> How can I help you today?',
model_id="kugel-3.5",
voice_id=1071,
language="en",
)const audio = await client.tts.generate({
text: 'Welcome to KugelAudio. <break time="400ms"/> How can I help you today?',
modelId: 'kugel-3.5',
voiceId: 1071,
language: 'en',
});curl -X POST https://api.kugelaudio.com/v1/tts/generate \
-H "Authorization: Bearer $KUGELAUDIO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"text": "Welcome to KugelAudio. <break time=\"400ms\"/> How can I help you today?",
"model_id": "kugel-3.5",
"voice_id": 1071,
"language": "en"
}' \
--output output.pcmBreak tags work the same in streaming: send them inline with your text; tags split across token boundaries are reassembled by the server's text buffer.
Notes & limits
- Breaks survive normalization. The text around a break is normalized
independently and the pause is re-inserted afterwards, so
normalize: trueand break tags compose. - A break needs a word before it. A break at the start of a text, or one
you stream on its own between two sentences, is dropped: the next sentence
already starts after a short pause. A request that holds only break tags has
nothing to speak and is answered like an empty text. Put the break after
the word it should follow,
"Thanks.<break time="1s"/> Next question.", not into a message of its own. - Model support:
kugel-3andkugel-3.5support break tags (see Models). Most legacy model IDs do not: on those, break tags are stripped and synthesis continues without the pause. Usekugel-3orkugel-3.5. - Punctuation first. For natural rhythm, commas/periods/ellipses are usually better; see Punctuation as pacing. Breaks are for deliberate silences of a specific length.
- Inside
<spell>blocks, don't place break tags. Use the spell tag'sgroupattribute for paced codes instead.