ONLINEAGENT_OPS 2026.Q3 HOME ARTICLES CRAFT RECORD BLOG MAP HUBS FAQ SEARCH
HOMETHE CRAFTCHEAT SHEETSELEVENLABS
CHEAT SHEET · ELEVENLABS · TEXT-TO-SPEECH · VOICE

ElevenLabs Prompting Cheat Sheet

There is no prompt box: the script is the prompt, and Eleven v3 takes audio tags instead of break tags.

READ4 min
WORDS632
SECTIONS5
SOURCES1
TYPECHEAT SHEET
CHECKED16 SEP 26

ElevenLabs' best-practices documentation covers pauses, emphasis and audio tags. This sheet quotes it, including the parts that come with a warning.

TL;DR — THE SHORT VERSION
  • There is no prompt field. The text you submit is the performance.
  • Eleven v3 uses inline audio tags such as [laughs], [whispers] and [sighs], and ElevenLabs says it does not support SSML break tags.
  • Other models take break tags for pauses. Use them sparingly: too many in one generation can make the output unstable.
  • Punctuation does real work. ElevenLabs says ellipses add pauses and weight, and capitals add emphasis.
  • A tag cannot override the voice. A whispering voice will not start shouting because of a tag; test tags on the voice you actually use.
VENDOR-DOCUMENTED

Every rule on this sheet is quoted from the vendor's own published documentation, linked where it is used.

The formula

Script + punctuation + (tags that match your model)

ElevenLabs, Best practices, read at source 16 Sep 2026.

Template:

[Sentence with ordinary punctuation for rhythm.] [Ellipsis for a weighted pause …] [CAPITALS for emphasis.] [v3: inline audio tag] / [break-tag models: <break time="1.0s" />]

Worked example:

It was a VERY long day [sigh] … nobody listens anymore.

Breakdown — what each part does

01

The script

Ordinary prose with deliberate punctuation. This carries most of the performance. The example above is ElevenLabs' own.

02

Punctuation

ElevenLabs: “Standard punctuation provides natural speech rhythm”

03

Capitalisation

ElevenLabs: “Capitalization increases emphasis”

04

v3 audio tags

Inline square brackets for emotion and delivery. ElevenLabs: “The voice you choose and its training samples will affect tag effectiveness.”

05

Break tags (not v3)

<break time="x.xs" />, up to 3 seconds. Dashes and ellipses also pause, but ElevenLabs adds: “However, these are less consistent.”

WHICH SYNTAX FOR WHICH MODEL
Know which model you are on, then use that side only.
ELEVEN V3
Yes: Inline audio tags such as [laughs] or [whispers]
No: SSML break tags — not supported
Yes: Punctuation and capitals for rhythm and emphasis
Note: Professional Voice Clones may sound worse than on earlier models
BREAK-TAG MODELS
Yes: <break time="x.xs" /> for pauses up to 3 seconds
Note: Too many break tags can cause instability
Yes: Punctuation and capitals for rhythm and emphasis
ElevenLabs, Best practices, read at source 16 Sep 2026: “Eleven v3 does not support SSML break tags.” and “Use <break time="x.xs" /> for natural pauses up to 3 seconds.”

Does it take a negative prompt?

Not applicable — there is no prompt or negative field.NO

The nearest thing to a negative prompt here is restraint: fewer tags, not more.

ElevenLabs, Best practices, read at source 16 Sep 2026: “Using too many break tags in a single generation can cause instability. The AI might speed up, or introduce additional noises or audio artifacts.”

TAKEAWAY

Write the script with punctuation first. Add a tag only where the plain reading came out wrong.

What not to do

Do not use SSML break tags on Eleven v3. ElevenLabs says v3 does not support them.
Do not stack break tags. Too many can make the voice speed up or add artefacts.
Do not expect a tag to contradict the voice's training.
Do not assume Professional Voice Clones behave the same on v3. ElevenLabs: “Professional Voice Clones (PVCs) are currently not fully optimized for Eleven v3, resulting in potentially lower clone quality compared to earlier models.”
ORDER OF OPERATIONS

This is this site's working order, not an ElevenLabs rule: write the script with punctuation, generate, then add tags only where the plain reading failed. Tagging every line up front is the voice equivalent of a bloated negative prompt. An earlier version of this sheet said break tags on v3 "fail silently"; ElevenLabs says only that v3 does not support them, so that wording was removed.

Free or paid? What the free plan allows, who owns the output, and when you need to pay: free vs paid AI image, video and music tools.

Where this model sits among the others: which models take a negative prompt, and what each vendor publishes as its formula.

Voice and music prompting more generally is on prompting music and voice.

The checklist

Before you send it:

1 — Know which model you are on — v3 or a break-tag model?
2 — Using the syntax for that model only?
3 — Punctuation doing the work before any tags?
4 — Break tags few, and 3 seconds or less?
5 — Tags tested on the actual voice?
SOURCES

ElevenLabs, Text to Speech best practices All read at source on 16 Sep 2026.

Model versions and vendor documentation both move. Re-read the source before relying on a specific number.

Every other sheet in this set: all model cheat sheets.

ABOUTMETHODVERIFYPRIVACYCONTACTINDEXAI PROMPT GENEER · EVERY ARTICLE CARRIES ITS OWN CHECKED DATE