elevenlabs

Name: elevenlabs
Author: calesthio

320installs4.9kforks

How to install elevenlabs?

calesthio/openmontage/elevenlabs

$npx -y skills add calesthio/OpenMontage --skill elevenlabs

Installs into the current project.

Summary

Generate AI voiceovers, sound effects, and music using ElevenLabs APIs. Use when creating audio content for videos, podcasts, or games. Triggers include generating voiceovers, narration, dialogue, sound effects from descriptions, background music, soundtrack generation, voice clo

Files · 1

View on GitHub

SKILL.md

1	# ElevenLabs Audio Generation
2
3	Requires `ELEVENLABS_API_KEY` in `.env`.
4
5	## Text-to-Speech
6
7	```python
8	from elevenlabs.client import ElevenLabs
9	from elevenlabs import save, VoiceSettings
10	import os
11
12	client = ElevenLabs(api_key=os.getenv("ELEVENLABS_API_KEY"))
13
14	audio = client.text_to_speech.convert(
15	text="Welcome to my video!",
16	voice_id="JBFqnCBsd6RMkjVDRZzb",
17	model_id="eleven_multilingual_v2",
18	voice_settings=VoiceSettings(
19	stability=0.5,
20	similarity_boost=0.75,
21	style=0.5,
22	speed=1.0
23	)
24	)
25	save(audio, "voiceover.mp3")
26	```
27
28	### Models
29
30	\| Model \| Quality \| SSML Support \| Notes \|
31	\|-------\|---------\|--------------\|-------\|
32	\| `eleven_multilingual_v2` \| Highest consistency \| None \| Stable, production-ready, 29 languages \|
33	\| `eleven_flash_v2_5` \| Good \| `<break>`, `<phoneme>` \| Fast, supports pause/pronunciation tags \|
34	\| `eleven_turbo_v2_5` \| Good \| `<break>`, `<phoneme>` \| Fastest latency \|
35	\| `eleven_v3` \| Most expressive \| None \| Alpha — unreliable, needs prompt engineering \|
36
37	Choose: multilingual_v2 for reliability, flash/turbo for SSML control, v3 for maximum expressiveness (expect retakes).
38
39	### Voice Settings by Style
40
41	\| Style \| stability \| similarity \| style \| speed \|
42	\|-------\|-----------\|------------\|-------\|-------\|
43	\| Natural/professional \| 0.75-0.85 \| 0.9 \| 0.0-0.1 \| 1.0 \|
44	\| Conversational \| 0.5-0.6 \| 0.85 \| 0.3-0.4 \| 0.9-1.0 \|
45	\| Energetic/YouTuber \| 0.3-0.5 \| 0.75 \| 0.5-0.7 \| 1.0-1.1 \|
46
47	### Pauses Between Sections
48
49	With flash/turbo models: Use SSML break tags inline:
50	```
51	...end of section. <break time="1.5s" /> Start of next...
52	```
53	Max 3 seconds per break. Excessive breaks can cause speed artifacts.
54
55	With multilingual_v2 / v3: No SSML support. Options:
56	- Paragraph breaks (blank lines) — creates ~0.3-0.5s natural pause
57	- Post-process with ffmpeg: split audio and insert silence
58
59	WARNING: `...` (ellipsis) is NOT a reliable pause — it can be vocalized as a word/sound. Do not use ellipsis as a pause mechanism.
60
61	### Pronunciation Control
62
63	Phonetic spelling (any model): Write words as you want them pronounced:
64	- `Janus` → `Jan-us`
65	- `nginx` → `engine-x`
66	- Use dashes, capitals, apostrophes to guide pronunciation
67
68	SSML phoneme tags (flash/turbo only):
69	```
70	<phoneme alphabet="ipa" ph="ˈdʒeɪnəs">Janus</phoneme>
71	```
72
73	### Iterative Workflow
74
75	1. Generate → listen → identify pronunciation/pacing issues
76	2. Adjust: phonetic spellings, break tags, voice settings
77	3. Regenerate. If pauses aren't precise enough, add silence in post with ffmpeg rather than fighting the TTS engine.
78
79	## Voice Cloning
80
81	### Instant Voice Clone
82
83	```python
84	with open("sample.mp3", "rb") as f:
85	voice = client.voices.ivc.create(
86	name="My Voice",
87	files=[f],
88	remove_background_noise=True
89	)
90	print(f"Voice ID: {voice.voice_id}")
91	```
92
93	- Use `client.voices.ivc.create()` (not `client.voices.clone()`)
94	- Pass file handles in binary mode (`"rb"`), not paths
95	- Convert m4a first: `ffmpeg -i input.m4a -codec:a libmp3lame -qscale:a 2 output.mp3`
96	- Multiple samples (2-3 clips) improve accuracy
97	- Save voice ID for reuse
98
99	Professional Voice Clone: Requires Creator plan+, 30+ min audio. See [reference.md](reference.md).
100
101	## Sound Effects
102
103	Max 22 seconds per generation.
104
105	```python
106	result = client.text_to_sound_effects.convert(
107	text="Thunder rumbling followed by heavy rain",
108	duration_seconds=10,
109	prompt_influence=0.3
110	)
111	with open("thunder.mp3", "wb") as f:
112	for chunk in result:
113	f.write(chunk)
114	```
115
116	Prompt tips: Be specific — "Heavy footsteps on wooden floorboards, slow and deliberate, with creaking"
117
118	## Music Generation
119
120	10 seconds to 5 minutes. Use `client.music.compose()` (not `.generate()`).
121
122	```python
123	result = client.music.compose(
124	prompt="Upbeat indie rock, catchy guitar riff, energetic drums, travel vlog",
125	music_length_ms=60000,
126	force_instrumental=True
127	)
128	with open("music.mp3", "wb") as f:
129	for chunk in result:
130	f.write(chunk)
131	```
132
133	Prompt structure: Genre, mood, instruments, tempo, use case. Add "no vocals" or use `force_instrumental=True` for background music.
134
135	## Remotion Integration
136
137	### Complete Workflow: Script to Synchronized Scene
138
139	```
140	VOICEOVER-SCRIPT.md → voiceover.py → public/audio/ → Remotion composition
141	↓ ↓ ↓ ↓
142	Scene narration Generate MP3 Audio files <Audio> component
143	with durations per scene with timing synced to scenes
144	```
145
146	### Step 1: Generate Per-Scene Audio
147
148	Use the toolkit's voiceover tool to generate audio for each scene:
149
150	```bash
151	# Generate voiceover files for each sc

People who liked this also viewed

6 picks

Type

elevenlabs

320installs4.9kforks

View on GitHub

How to install elevenlabs?

calesthio/openmontage/elevenlabs

$npx -y skills add calesthio/OpenMontage --skill elevenlabs

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# ElevenLabs Audio Generation
2
3	Requires `ELEVENLABS_API_KEY` in `.env`.
4
5	## Text-to-Speech
6
7	```python
8	from elevenlabs.client import ElevenLabs
9	from elevenlabs import save, VoiceSettings
10	import os
11
12	client = ElevenLabs(api_key=os.getenv("ELEVENLABS_API_KEY"))
13
14	audio = client.text_to_speech.convert(
15	text="Welcome to my video!",
16	voice_id="JBFqnCBsd6RMkjVDRZzb",
17	model_id="eleven_multilingual_v2",
18	voice_settings=VoiceSettings(
19	stability=0.5,
20	similarity_boost=0.75,
21	style=0.5,
22	speed=1.0
23	)
24	)
25	save(audio, "voiceover.mp3")
26	```
27
28	### Models
29
30	\| Model \| Quality \| SSML Support \| Notes \|
31	\|-------\|---------\|--------------\|-------\|
32	\| `eleven_multilingual_v2` \| Highest consistency \| None \| Stable, production-ready, 29 languages \|
33	\| `eleven_flash_v2_5` \| Good \| `<break>`, `<phoneme>` \| Fast, supports pause/pronunciation tags \|
34	\| `eleven_turbo_v2_5` \| Good \| `<break>`, `<phoneme>` \| Fastest latency \|
35	\| `eleven_v3` \| Most expressive \| None \| Alpha — unreliable, needs prompt engineering \|
36
37	Choose: multilingual_v2 for reliability, flash/turbo for SSML control, v3 for maximum expressiveness (expect retakes).
38
39	### Voice Settings by Style
40
41	\| Style \| stability \| similarity \| style \| speed \|
42	\|-------\|-----------\|------------\|-------\|-------\|
43	\| Natural/professional \| 0.75-0.85 \| 0.9 \| 0.0-0.1 \| 1.0 \|
44	\| Conversational \| 0.5-0.6 \| 0.85 \| 0.3-0.4 \| 0.9-1.0 \|
45	\| Energetic/YouTuber \| 0.3-0.5 \| 0.75 \| 0.5-0.7 \| 1.0-1.1 \|
46
47	### Pauses Between Sections
48
49	With flash/turbo models: Use SSML break tags inline:
50	```
51	...end of section. <break time="1.5s" /> Start of next...
52	```
53	Max 3 seconds per break. Excessive breaks can cause speed artifacts.
54
55	With multilingual_v2 / v3: No SSML support. Options:
56	- Paragraph breaks (blank lines) — creates ~0.3-0.5s natural pause
57	- Post-process with ffmpeg: split audio and insert silence
58
59	WARNING: `...` (ellipsis) is NOT a reliable pause — it can be vocalized as a word/sound. Do not use ellipsis as a pause mechanism.
60
61	### Pronunciation Control
62
63	Phonetic spelling (any model): Write words as you want them pronounced:
64	- `Janus` → `Jan-us`
65	- `nginx` → `engine-x`
66	- Use dashes, capitals, apostrophes to guide pronunciation
67
68	SSML phoneme tags (flash/turbo only):
69	```
70	<phoneme alphabet="ipa" ph="ˈdʒeɪnəs">Janus</phoneme>
71	```
72
73	### Iterative Workflow
74
75	1. Generate → listen → identify pronunciation/pacing issues
76	2. Adjust: phonetic spellings, break tags, voice settings
77	3. Regenerate. If pauses aren't precise enough, add silence in post with ffmpeg rather than fighting the TTS engine.
78
79	## Voice Cloning
80
81	### Instant Voice Clone
82
83	```python
84	with open("sample.mp3", "rb") as f:
85	voice = client.voices.ivc.create(
86	name="My Voice",
87	files=[f],
88	remove_background_noise=True
89	)
90	print(f"Voice ID: {voice.voice_id}")
91	```
92
93	- Use `client.voices.ivc.create()` (not `client.voices.clone()`)
94	- Pass file handles in binary mode (`"rb"`), not paths
95	- Convert m4a first: `ffmpeg -i input.m4a -codec:a libmp3lame -qscale:a 2 output.mp3`
96	- Multiple samples (2-3 clips) improve accuracy
97	- Save voice ID for reuse
98
99	Professional Voice Clone: Requires Creator plan+, 30+ min audio. See [reference.md](reference.md).
100
101	## Sound Effects
102
103	Max 22 seconds per generation.
104
105	```python
106	result = client.text_to_sound_effects.convert(
107	text="Thunder rumbling followed by heavy rain",
108	duration_seconds=10,
109	prompt_influence=0.3
110	)
111	with open("thunder.mp3", "wb") as f:
112	for chunk in result:
113	f.write(chunk)
114	```
115
116	Prompt tips: Be specific — "Heavy footsteps on wooden floorboards, slow and deliberate, with creaking"
117
118	## Music Generation
119
120	10 seconds to 5 minutes. Use `client.music.compose()` (not `.generate()`).
121
122	```python
123	result = client.music.compose(
124	prompt="Upbeat indie rock, catchy guitar riff, energetic drums, travel vlog",
125	music_length_ms=60000,
126	force_instrumental=True
127	)
128	with open("music.mp3", "wb") as f:
129	for chunk in result:
130	f.write(chunk)
131	```
132
133	Prompt structure: Genre, mood, instruments, tempo, use case. Add "no vocals" or use `force_instrumental=True` for background music.
134
135	## Remotion Integration
136
137	### Complete Workflow: Script to Synchronized Scene
138
139	```
140	VOICEOVER-SCRIPT.md → voiceover.py → public/audio/ → Remotion composition
141	↓ ↓ ↓ ↓
142	Scene narration Generate MP3 Audio files <Audio> component
143	with durations per scene with timing synced to scenes
144	```
145
146	### Step 1: Generate Per-Scene Audio
147
148	Use the toolkit's voiceover tool to generate audio for each scene:
149
150	```bash
151	# Generate voiceover files for each sc

People who liked this also viewed

6 picks

Type