Multimodal
Audio Support
Transcribe and analyze audio files with audio-capable models
Features
- Transcription: Convert speech to text
- Audio Analysis: Summarize meetings, extract topics
- Multiple Formats: MP3, WAV
- Flexible Input: Local files, remote URLs, or base64 data
- Easy Upload: Drag & drop in the web UI
Using Audio in CLI
# Transcribe audio using default template
llms --audio ./recording.mp3
# Local audio file with prompt
llms --audio ./meeting.wav "Summarize this meeting recording"
# Remote audio URL
llms --audio https://example.org/podcast.mp3 "What are the key points?"
# With specific audio model
llms -m gpt-4o-audio-preview --audio interview.mp3 "Extract main topics"
# Combined with system prompt
llms -s "You're a transcription specialist" --audio talk.mp3 "Provide transcript"Using Audio in UI
Drag and drop audio files or use the attach button to upload.
Audio-Capable Models
Models that support audio processing:
- OpenAI: gpt-4o-audio-preview
- Google: gemini-2.5-pro, gemini-2.5-flash, gemini-2.5-flash-lite
Custom Audio Template
{
"model": "gpt-4o-audio-preview",
"messages": [
{
"role": "user",
"content": [
{
"type": "input_audio",
"input_audio": {
"data": "",
"format": "mp3"
}
},
{
"type": "text",
"text": "Please transcribe this audio"
}
]
}
]
}llms --chat audio-request.json --audio speech.wavUse Cases
- Meeting Transcription: Convert meeting recordings to text
- Interview Analysis: Extract key points from interviews
- Podcast Summaries: Summarize podcast episodes
- Voice Notes: Transcribe voice memos
Tips for Best Results
- Use clear audio with minimal background noise
- Consider file size limits
- MP3 is typically smaller than WAV
Performance Considerations
- Larger files take longer to process
- Consider using lighter models for simple tasks
- Remote URLs may be slower due to download time