Scenario ElevenLabs Audio
Overview
ElevenLabs covers five audio lanes on Scenario: speech, music, sound effects, dubbing, and voice transforms. Eleven members ship side by side at authoring time, one job each: the deciding move is picking the right member, not coaxing one model into a mode. Discover them with
and treat
as the contract: the members agree on almost nothing.
Connection and the core loop: see the
skill in this repo; model-agnostic audio work: the
skill. If a sibling skill named here is missing from your available skills, ask the user to install it (
npx skills add scenario-labs/skills --skill <name>
); unattended, proceed from tool schemas and flag the gap.
Quick reference
| Job | Member | Deciding inputs |
|---|
| Expressive speech, dialog | Eleven v3 | with inline tags |
| Narration, pinned language | Multilingual 2 | |
| Fast or bulk speech | Turbo 2.5 | same shape, lower cost |
| Music from one prompt | Music v2 | , , |
| Structured song | Music Advanced v2 | required array |
| SFX and loops | Sound Effects 2 | , , |
| Localize, protect names | Dubbing v2 | , , |
| Localize, speaker control | Dubbing | , |
| Re-voice a recording | Speech to Speech, Voice Changer | (up to 5 min) plus a voice |
| Clean a noisy voice | Voice Isolator | |
The speaking members share one voice contract:
takes a cloned ElevenLabs voice and always wins;
picks a preset (21 at authoring time) and is ignored beside it; neither set means the Adam preset. Voice Changer is Speech to Speech plus delivery dials (
,
,
,
). WAV output exists on the TTS members alone;
elsewhere is MP3 or Opus, absent on Dubbing and Voice Isolator.
gives repeatability, except on Sound Effects 2, both Dubbing members, and Voice Isolator. Cost rides the content:
on TTS (40,000 characters at authoring time),
on Music v2,
on Advanced,
or
on the dubbing and voice members, so
long jobs and member comparisons.
Speech: tags on v3, dials elsewhere
Eleven v3 reads inline audio tags in the text, [whispers], [excited], [sighs], to steer delivery moment to moment across 70+ languages, and carries multi-speaker dialogue. Multilingual 2 and Turbo 2.5 do not read tags: direction there lives in
,
, and
(0.7 to 1.2), and
(ISO 639-1) pins the language, a field v3 lacks. Turbo trades expressiveness for cost, roughly half the other two per run at authoring time.
Music: one prompt or thirty sections
Music v2 takes one
(mood, genre, instruments, tempo),
(3 to 600 at authoring time, cost impact), and
to suppress vocals: asking in prose is unreliable. Music Advanced v2 requires
, up to 30 ordered segments at authoring time, each with
, its own
(3 to 120),
and
(up to 10 each), and
(high binds a segment to its neighbors, low frees it). Section grammar: square brackets label ([Verse], [Chorus]), curly braces direct ({soft piano intro}), and plain text is sung as lyrics. Advanced has no instrumental flag, so any plain text will be sung. Neither music member takes
: one run is one composition, and sung lyrics land differently on every draw, so a batch of takes is several parallel runs with distinct
values, each priced alone on its duration, which costs what a batch field would have and returns separate files with real boundaries. Two
blocks in one run to get two takes returns one file with no split point.
Dubbing replaces the track, not the lips
Both Dubbing members take an audio or video
with a required
, and the output follows the input kind: video in, dubbed video out. Neither re-animates lips. Source auto-detection is spelled differently:
on v2, an empty string on the older member. Pick v2 for
(names, brands, and jargon preserved verbatim through translation); pick the older Dubbing for
(0 auto-detects, up to 10),
,
, and
video. Their prices differ several-fold, so
both when either fits. Speech to Speech, Voice Changer and Voice Isolator, by contrast, take
only, whatever the catalog blurb says: for a clip, pull the track with
model_scenario-audio-extract
, run the member, then lay the result over the clip in
model_scenario-compose-video
, the clip as a video layer with
(fixed ids: each is Scenario's single deterministic tool for its operation, so discovery would only re-derive it). The same compositor lays a Music v2 bed under a video; layer contract in
.
Worked example: dub a trailer into Spanish
- with ,
query="elevenlabs dubbing"
model_elevenlabs-dubbing-v2
- with that id: fields and defaults before anything else.
- the trailer video (see the skill) to get an asset id.
- with that , , and the exact
parameters={"file": "asset_x", "targetLang": "es", "keyterms": ["Aetherfall", "Kestrel Squad"]}
- Repeat with , then with the returned job id, re-called with on timeout, never a second .
- the result (a video, because the input was) and confirm the keyterms survived; with no to save it.
Common mistakes
- Setting next to and expecting it to apply: a set always wins.
- Inline tags on Multilingual 2 or Turbo 2.5: that grammar is Eleven v3's; elsewhere a tag can be read aloud.
- Writing "instrumental" in a Music v2 prompt instead of setting ; on Music Advanced there is no flag and plain section text is always sung.
- Expecting lip-sync from Dubbing: the track changes, the picture does not.
- Reaching for on the music members, or repeating one long run hoping for a different take without changing : same seed and settings reproduce the same music.
- Carrying one member's caps to another: 40,000 characters, 600 seconds, 30 sections, and 5 minutes of input audio are each true of one lane and false of the next.