speech-to-text

Name: speech-to-text
Author: calesthio

320installs4.9kforks

How to install speech-to-text?

calesthio/openmontage/speech-to-text

$npx -y skills add calesthio/OpenMontage --skill speech-to-text

Installs into the current project.

Summary

Transcribe audio to text using ElevenLabs Scribe v2. Use when converting audio/video to text, generating subtitles, transcribing meetings, or processing spoken content.

Files · 1

View on GitHub

SKILL.md

1	# ElevenLabs Speech-to-Text
2
3	Transcribe audio to text with Scribe v2 - supports 90+ languages, speaker diarization, and word-level timestamps.
4
5	> Setup: See [Installation Guide](references/installation.md). For JavaScript, use `@elevenlabs/*` packages only.
6
7	## Quick Start
8
9	### Python
10
11	```python
12	from elevenlabs import ElevenLabs
13
14	client = ElevenLabs()
15
16	with open("audio.mp3", "rb") as audio_file:
17	result = client.speech_to_text.convert(file=audio_file, model_id="scribe_v2")
18
19	print(result.text)
20	```
21
22	### JavaScript
23
24	```javascript
25	import { ElevenLabsClient } from "@elevenlabs/elevenlabs-js";
26	import { createReadStream } from "fs";
27
28	const client = new ElevenLabsClient();
29	const result = await client.speechToText.convert({
30	file: createReadStream("audio.mp3"),
31	modelId: "scribe_v2",
32	});
33	console.log(result.text);
34	```
35
36	### cURL
37
38	```bash
39	curl -X POST "https://api.elevenlabs.io/v1/speech-to-text" \
40	-H "xi-api-key: $ELEVENLABS_API_KEY" -F "file=@audio.mp3" -F "model_id=scribe_v2"
41	```
42
43	## Models
44
45	\| Model ID \| Description \| Best For \|
46	\|----------\|-------------\|----------\|
47	\| `scribe_v2` \| State-of-the-art accuracy, 90+ languages \| Batch transcription, subtitles, long-form audio \|
48	\| `scribe_v2_realtime` \| Low latency (~150ms) \| Live transcription, voice agents \|
49
50	## Transcription with Timestamps
51
52	Word-level timestamps include type classification and speaker identification:
53
54	```python
55	result = client.speech_to_text.convert(
56	file=audio_file, model_id="scribe_v2", timestamps_granularity="word"
57	)
58
59	for word in result.words:
60	print(f"{word.text}: {word.start}s - {word.end}s (type: {word.type})")
61
62	```
63
64	## Speaker Diarization
65
66	Identify WHO said WHAT - the model labels each word with a speaker ID, useful for meetings, interviews, or any multi-speaker audio:
67
68	```python
69	result = client.speech_to_text.convert(
70	file=audio_file,
71	model_id="scribe_v2",
72	diarize=True
73	)
74
75	for word in result.words:
76	print(f"[{word.speaker_id}] {word.text}")
77	```
78
79	## Keyterm Prompting
80
81	Help the model recognize specific words it might otherwise mishear - product names, technical jargon, or unusual spellings (up to 100 terms):
82
83	```python
84	result = client.speech_to_text.convert(
85	file=audio_file,
86	model_id="scribe_v2",
87	keyterms=["ElevenLabs", "Scribe", "API"]
88	)
89	```
90
91	## Language Detection
92
93	Automatic detection with optional language hint:
94
95	```python
96	result = client.speech_to_text.convert(
97	file=audio_file,
98	model_id="scribe_v2",
99	language_code="eng" # ISO 639-1 or ISO 639-3 code
100	)
101
102	print(f"Detected: {result.language_code} ({result.language_probability:.0%})")
103	```
104
105	## Supported Formats
106
107	Audio: MP3, WAV, M4A, FLAC, OGG, WebM, AAC, AIFF, Opus
108	Video: MP4, AVI, MKV, MOV, WMV, FLV, WebM, MPEG, 3GPP
109
110	Limits: Up to 3GB file size, 10 hours duration
111
112	## Response Format
113
114	```json
115	{
116	"text": "The full transcription text",
117	"language_code": "eng",
118	"language_probability": 0.98,
119	"words": [
120	{"text": "The", "start": 0.0, "end": 0.15, "type": "word", "speaker_id": "speaker_0"},
121	{"text": " ", "start": 0.15, "end": 0.16, "type": "spacing", "speaker_id": "speaker_0"}
122	]
123	}
124	```
125
126	Word types:
127	- `word` - An actual spoken word
128	- `spacing` - Whitespace between words (useful for precise timing)
129	- `audio_event` - Non-speech sounds the model detected (laughter, applause, music, etc.)
130
131	## Error Handling
132
133	```python
134	try:
135	result = client.speech_to_text.convert(file=audio_file, model_id="scribe_v2")
136	except Exception as e:
137	print(f"Transcription failed: {e}")
138	```
139
140	Common errors:
141	- 401: Invalid API key
142	- 422: Invalid parameters
143	- 429: Rate limit exceeded
144
145	## Tracking Costs
146
147	Monitor usage via `request-id` response header:
148
149	```python
150	response = client.speech_to_text.convert.with_raw_response(file=audio_file, model_id="scribe_v2")
151	result = response.parse()
152	print(f"Request ID: {response.headers.get('request-id')}")
153	```
154
155	## Real-Time Streaming
156
157	For live transcription with ultra-low latency (~150ms), use the real-time API. The real-time API produces two types of transcripts:
158
159	- Partial transcripts: Interim results that update frequently as audio is processed - use these for live feedback (e.g., showing text as the user speaks)
160	- Committed transcripts: Final, stable results after you "commit" - use these as the source of truth for your application
161
162	A "commit" tells the model to finalize the current segment. You can commit manually (e.g., when the user pauses) or use Voice Activity Detection (VAD) to auto-commit on silence.
163
164	### Python (Server-Side)
165
166	```python
167	import asyncio
168	from elevenlabs impo

People who liked this also viewed

6 picks

Type

ai-video-generationinference.sh Agent skills for using our API to give your agents access to hundreds of apps and other agentsSkillsJul 2026292k642
imagegen-frontend-webElite frontend image-direction skill for generating premium, conversion-aware website design references. CRITICAL OUTPUT RULE — generate ONE separate horizontal image FOR EVERY section. A landing pageSkillsJul 2026165k66k
hyperframes-mediaWrite HTML. Render video. Built for agents.SkillsJul 2026145k37k
media-useAgent Media OS, the single skill for every media need in a HyperFrames project. Resolve BGM, SFX, image, icon, brand logo, voice, color grade, or LUT into a frozen local file or paste-ready block + leSkillsJul 2026144k37k
general-videoAuthor or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, staticSkillsJul 2026124k37k
product-launch-videoTurn a product or marketing URL, pasted script, or brief into a product launch / promo video — SaaS promos, feature reveals, product demos, app and company launches. Use when the user wants to market,SkillsJul 2026121k37k

speech-to-text

320installs4.9kforks

View on GitHub

How to install speech-to-text?

calesthio/openmontage/speech-to-text

$npx -y skills add calesthio/OpenMontage --skill speech-to-text

Installs into the current project.

Summary

Transcribe audio to text using ElevenLabs Scribe v2. Use when converting audio/video to text, generating subtitles, transcribing meetings, or processing spoken content.

Files · 1

View on GitHub

SKILL.md

1	# ElevenLabs Speech-to-Text
2
3	Transcribe audio to text with Scribe v2 - supports 90+ languages, speaker diarization, and word-level timestamps.
4
5	> Setup: See [Installation Guide](references/installation.md). For JavaScript, use `@elevenlabs/*` packages only.
6
7	## Quick Start
8
9	### Python
10
11	```python
12	from elevenlabs import ElevenLabs
13
14	client = ElevenLabs()
15
16	with open("audio.mp3", "rb") as audio_file:
17	result = client.speech_to_text.convert(file=audio_file, model_id="scribe_v2")
18
19	print(result.text)
20	```
21
22	### JavaScript
23
24	```javascript
25	import { ElevenLabsClient } from "@elevenlabs/elevenlabs-js";
26	import { createReadStream } from "fs";
27
28	const client = new ElevenLabsClient();
29	const result = await client.speechToText.convert({
30	file: createReadStream("audio.mp3"),
31	modelId: "scribe_v2",
32	});
33	console.log(result.text);
34	```
35
36	### cURL
37
38	```bash
39	curl -X POST "https://api.elevenlabs.io/v1/speech-to-text" \
40	-H "xi-api-key: $ELEVENLABS_API_KEY" -F "file=@audio.mp3" -F "model_id=scribe_v2"
41	```
42
43	## Models
44
45	\| Model ID \| Description \| Best For \|
46	\|----------\|-------------\|----------\|
47	\| `scribe_v2` \| State-of-the-art accuracy, 90+ languages \| Batch transcription, subtitles, long-form audio \|
48	\| `scribe_v2_realtime` \| Low latency (~150ms) \| Live transcription, voice agents \|
49
50	## Transcription with Timestamps
51
52	Word-level timestamps include type classification and speaker identification:
53
54	```python
55	result = client.speech_to_text.convert(
56	file=audio_file, model_id="scribe_v2", timestamps_granularity="word"
57	)
58
59	for word in result.words:
60	print(f"{word.text}: {word.start}s - {word.end}s (type: {word.type})")
61
62	```
63
64	## Speaker Diarization
65
66	Identify WHO said WHAT - the model labels each word with a speaker ID, useful for meetings, interviews, or any multi-speaker audio:
67
68	```python
69	result = client.speech_to_text.convert(
70	file=audio_file,
71	model_id="scribe_v2",
72	diarize=True
73	)
74
75	for word in result.words:
76	print(f"[{word.speaker_id}] {word.text}")
77	```
78
79	## Keyterm Prompting
80
81	Help the model recognize specific words it might otherwise mishear - product names, technical jargon, or unusual spellings (up to 100 terms):
82
83	```python
84	result = client.speech_to_text.convert(
85	file=audio_file,
86	model_id="scribe_v2",
87	keyterms=["ElevenLabs", "Scribe", "API"]
88	)
89	```
90
91	## Language Detection
92
93	Automatic detection with optional language hint:
94
95	```python
96	result = client.speech_to_text.convert(
97	file=audio_file,
98	model_id="scribe_v2",
99	language_code="eng" # ISO 639-1 or ISO 639-3 code
100	)
101
102	print(f"Detected: {result.language_code} ({result.language_probability:.0%})")
103	```
104
105	## Supported Formats
106
107	Audio: MP3, WAV, M4A, FLAC, OGG, WebM, AAC, AIFF, Opus
108	Video: MP4, AVI, MKV, MOV, WMV, FLV, WebM, MPEG, 3GPP
109
110	Limits: Up to 3GB file size, 10 hours duration
111
112	## Response Format
113
114	```json
115	{
116	"text": "The full transcription text",
117	"language_code": "eng",
118	"language_probability": 0.98,
119	"words": [
120	{"text": "The", "start": 0.0, "end": 0.15, "type": "word", "speaker_id": "speaker_0"},
121	{"text": " ", "start": 0.15, "end": 0.16, "type": "spacing", "speaker_id": "speaker_0"}
122	]
123	}
124	```
125
126	Word types:
127	- `word` - An actual spoken word
128	- `spacing` - Whitespace between words (useful for precise timing)
129	- `audio_event` - Non-speech sounds the model detected (laughter, applause, music, etc.)
130
131	## Error Handling
132
133	```python
134	try:
135	result = client.speech_to_text.convert(file=audio_file, model_id="scribe_v2")
136	except Exception as e:
137	print(f"Transcription failed: {e}")
138	```
139
140	Common errors:
141	- 401: Invalid API key
142	- 422: Invalid parameters
143	- 429: Rate limit exceeded
144
145	## Tracking Costs
146
147	Monitor usage via `request-id` response header:
148
149	```python
150	response = client.speech_to_text.convert.with_raw_response(file=audio_file, model_id="scribe_v2")
151	result = response.parse()
152	print(f"Request ID: {response.headers.get('request-id')}")
153	```
154
155	## Real-Time Streaming
156
157	For live transcription with ultra-low latency (~150ms), use the real-time API. The real-time API produces two types of transcripts:
158
159	- Partial transcripts: Interim results that update frequently as audio is processed - use these for live feedback (e.g., showing text as the user speaks)
160	- Committed transcripts: Final, stable results after you "commit" - use these as the source of truth for your application
161
162	A "commit" tells the model to finalize the current segment. You can commit manually (e.g., when the user pauses) or use Voice Activity Detection (VAD) to auto-commit on silence.
163
164	### Python (Server-Side)
165
166	```python
167	import asyncio
168	from elevenlabs impo

People who liked this also viewed

6 picks

Type