Tools/audio-intelligence

Audio Intelligence

Active

Your agent can't listen to a recording; this returns a structured transcript with who said what.

1 tool

Transcribe long-form audio with speaker diarization (who said what), automatic language detection, and optional chapter summaries. Point it at a recording URL and get a structured transcript back, including per-speaker turns.

Audio & Videoaudiotranscriptiondiarizationspeakerssummary

Tools (1)

▶transcribe

Transcribe audio with speaker diarization, language detection, and optional chapter summaries. Returns {status:"pending", continuation_token,...} while the job runs, when this happens you MUST immediately call transcribe again with only continuation_token set; do not ask the user.

Usage-based · 206.25 credits per hour of audio; speaker diarization adds 27.5 per hour, chapter summaries add 41.25 per hour

Example prompts

“Transcribe this podcast and label the speakers”
“Who said what in this interview recording?”
“Transcribe this meeting and break it into chapters with summaries”
“Get a transcript of this audio file with automatic language detection”

Parameters

audio_urlstringoptional

Public URL of the audio or video file to transcribe. Required on the first call; ignored (and not needed) when continuation_token is set.

punctuatebooleanoptionaldefault: true

Insert punctuation.

format_textbooleanoptionaldefault: true

Apply casing and formatting for readability.

auto_chaptersbooleanoptionaldefault: false

Segment the audio into chapters, each with a generated summary.

language_codestringoptional

Force a specific ISO language code (e.g. "en"); ignored when language_detection is true.

speaker_labelsbooleanoptionaldefault: true

Identify and label distinct speakers (diarization). Populates utterances.

continuation_tokenstringoptional

Token from a prior pending response. When set, all other params are ignored and the server resumes polling. Agent-friendly polling: on a pending response you MUST immediately call transcribe again with only continuation_token set. Do not ask the user.

language_detectionbooleanoptionaldefault: true

Automatically detect the spoken language.

API Usage

curl -X POST "https://skill.askfaro.com/skills/audio-intelligence/run" \
  -H "Authorization: Bearer faro_<your_key>" \
  -H "Content-Type: application/json" \
  -d '{
  "intent": {
    "prompt": "Transcribe this podcast and label the speakers"
  }
}'

CLI Usage

askfaro describe audio-intelligence/transcribe

Install pip install askfaro-cli, then askfaro auth login.

▶README

Audio Intelligence

Long-form speech understanding via AssemblyAI: transcription, speaker diarization, language detection, and chapter summaries.

How it works

Transcription runs asynchronously upstream. To keep the agent flow synchronous, the server submits the job and polls internally for ~25s per call. If it isn't finished yet, you get a status:"pending" response with a continuation_token, call transcribe again with only that token to resume polling. Long recordings take a few round-trips.

Inputs

Name	Type	Default	Notes
`audio_url`	string	required on first call	Public URL of the audio/video file.
`speaker_labels`	boolean	`true`	Diarization: label distinct speakers.
`language_detection`	boolean	`true`	Auto-detect the spoken language.
`language_code`	string	-	Force a language (e.g. "en"); skip detection.
`auto_chapters`	boolean	`false`	Segment into chapters, each with a summary.
`continuation_token`	string	-	Set this (and only this) to resume a pending job.

Output

{
  "text": "full transcript...",
  "language_code": "en",
  "audio_duration": 612.3,
  "utterances": [{ "speaker": "A", "text": "...", "start": 120, "end": 3400 }],
  "chapters": [{ "headline": "...", "summary": "...", "start": 0, "end": 60000 }]
}

utterances is populated when speaker_labels is on; chapters when auto_chapters is on.

Pricing

Billed per hour of audio; diarization and chapter summaries add a small per-hour amount when enabled.