Back to Skills

Cloud Transcription (Whisper API)

Convert calls and voice recordings to text for instant meeting notes and searchable records

AI ServiceActive

What It Does

Cloud Transcription uses OpenAI's Whisper API to convert audio from meetings, calls, and voice memos into accurate text transcripts. Whether you're recording a Zoom call or dictating notes, this skill delivers high-quality transcriptions in over 50 languages.

Voice to TextMulti-languageMeeting NotesCall RecordingWhisper API

In a Nutshell

🎤
Audio Upload accept audio files up to 25MB
🌍
Multi-language transcribe 50+ languages automatically
Fast Processing get transcripts in seconds
📝
Meeting Notes convert calls to searchable text
🔍
Speaker Detection identify different speakers (when available)

Use Cases

Meeting Documentation

Record and transcribe team meetings for instant searchable notes

Interview Analysis

Transcribe customer interviews and research calls

Voice Memos

Dictate ideas and get them transcribed on the spot

Podcast Processing

Create show notes and transcripts for audio content

How to Use

Step 1

Upload or record audio

Send an audio file or start a live recording through your preferred channel.

Supported formats: MP3, M4A, WAV, WEBM, MP4

Step 2

Automatic processing

The assistant sends audio to Whisper API and retrieves the transcript automatically.

Step 3

Receive formatted transcript

Get clean text output with timestamps and optional speaker labels.

Step 4

Store or share

Save to Notion, send via email, or export to your preferred format.

Command Examples

You say:

Transcribe this meeting recording [audio file]

Assistant responds:

[00:00] John: Let's start with Q4 goals. [00:15] Sarah: We need to focus on retention...

You say:

Convert my voice memo to text

Assistant responds:

Transcribed: "Remember to follow up with the client about the proposal. Send revised pricing by Thursday..."

You say:

Transcribe the last 10 minutes of this call

Assistant responds:

Transcript ready (10:34 duration): Discussed action items, assigned owners, set next meeting for Dec 15.

Limits & Behavior

ParameterLimitNotes
File size25 MBcompress large files before upload
Duration3 hourssplit longer recordings
Daily requests500 filesPro plan unlimited
Concurrent jobs5 at oncequeues additional files

Models & Modes

ModelSpeedAccuracyBest For
Whisper LargeMediumHighestcritical transcripts, multi-speaker
Whisper MediumFastHighgeneral meetings and calls
Whisper SmallVery FastGoodquick voice memos

FAQ

Setup Requirements

OpenAI API key configured
Audio recording capability
Internet connection for API calls
File upload permissions

Troubleshooting

ErrorMeaningAction
FILE_TOO_LARGEExceeds 25MB limitCompress audio or split file
UNSUPPORTED_FORMATAudio format not recognizedConvert to MP3, M4A, or WAV
API_TIMEOUTProcessing took too longRetry with shorter clip
LOW_QUALITYPoor transcription resultRe-record with better audio quality