> ## Documentation Index
> Fetch the complete documentation index at: https://docs.famulor.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Prompting for speech

> Write numbers, dates, and identifiers so they come out sounding right when your assistant reads them aloud

Your prompt doesn't just decide what the assistant says — it decides how spoken details land. A phone number typed as digits and a phone number spelled out for speech can sound very different once a voice reads them aloud. A few habits make the difference.

<Note>
  This page is about how you *write* the words in your prompt and knowledge base. To fix how a specific name or term is pronounced everywhere, use the [pronunciation dictionary](/assistants/models-and-voices#speaking-style) instead — the two are complementary.
</Note>

## Write the spoken form directly

The most reliable technique is also the simplest: instead of relying on the model to reformat digits on the fly, give it the exact words to say.

**Phone numbers** — spell out each digit and group them for clarity:

```text theme={null}
When asked for the callback number, say:
"five five five, zero one three, four" —
not "5550134."
```

**Email addresses** — spell every character, and say the symbols as words:

```text theme={null}
Spell email addresses letter by letter. The @ is "at" and the . is "dot."
Example: name@company.com → "n, a, m, e, at, company, dot, com."
```

**Websites** — say recognizable words normally, and spell out any segment made only of letters:

```text theme={null}
Pronounce real words normally and spell out abbreviations letter by
letter. Always say "dot" before the top-level domain:
"yourcompany.io" → "your company, dot, i, o."
```

**Times and dates** — use a spoken format, never a raw timestamp:

```text theme={null}
Say times in words: "1:00 PM" → "one PM," "15:30" → "three thirty PM."
Never read a date as digits — "07/05" means two different days
depending on who's listening. Say "the fifth of July" instead.
```

## Pace long strings of digits

Long numbers — account numbers, IBANs, confirmation codes — are easier to follow when they're broken into short groups with a natural pause between them, rather than read as one unbroken string. A few short groups, separated by commas, are usually enough. Keep the grouping consistent within a prompt, and listen to the result before relying on it — pacing on punctuation can vary slightly by voice.

<Tip>
  Don't overdo it. Long chains of pauses make speech sound unnatural — reserve them for information the caller genuinely needs a moment to write down.
</Tip>

## Giving the caller a moment

If a caller says they need a second — to grab an account number, for example — tell the assistant explicitly how to respond rather than leaving it to guess:

```text theme={null}
If the caller says "one moment," "hold on," or similar, reply briefly
("Take your time") and then wait for them to continue. Don't fill the
silence with another question.
```

## Test what it actually sounds like

Wording that reads fine on screen doesn't always sound right out loud. Use the assistant editor's **Test** button to place a real voice call and listen — especially after changing a voice, since pacing and pronunciation both depend on it.

<Tip>
  See [Prompt writing](/assistants/prompt-writing) for structuring the rest of the prompt, and [Example prompts](/assistants/example-prompts/overview) for ready-to-adapt templates.
</Tip>