Voxtral Speech Recognition Demo
Experience the power of Voxtral's open-source speech understanding AI. Upload audio or video files, record live speech, or try our sample content to see Voxtral's superior transcription and multilingual capabilities in action.
Input Audio
Upload Audio File
Drag and drop audio or video files here or click to browse
Supported formats: MP3, WAV, M4A, FLAC, MP4, AVI, MOV (max 50MB)
Record Live Audio
Try Sample Content
Voxtral Results
Upload or record audio to see Voxtral's transcription results here.
Ask a question about the audio content to see Voxtral's understanding capabilities.
Upload audio to see Voxtral's automatic summarization capabilities.
Export Results
Voxtral Demo Features
Experience the full range of Voxtral's capabilities through our interactive demonstration.
High Accuracy Transcription
State-of-the-art word error rate (WER) that outperforms Whisper and other leading models.
Multilingual Support
Automatic language detection and superior performance across 8+ languages including English, Spanish, French, Portuguese, Hindi, German, Dutch, and Italian.
Direct Q&A
Ask questions directly about audio content without additional LLM chaining. Voxtral understands context and provides relevant answers.
Automatic Summarization
Generate concise summaries of long audio content automatically. Perfect for podcasts, lectures, and meetings.
Real-time Processing
Experience live transcription and processing capabilities with minimal latency for interactive applications.
Open Source
All Voxtral models are open source under Apache 2.0 license. Deploy locally, customize, and build without vendor lock-in.
Performance Comparison
See how Voxtral compares to other speech recognition solutions in real-world scenarios.
| Feature | Voxtral Small | Voxtral Mini | Whisper Large v3 | GPT-4o Mini |
|---|---|---|---|---|
| English WER | SOTA | Excellent | Good | Good |
| Multilingual Performance | Superior | Superior | Good | Good |
| API Cost (per minute) | $0.001 | $0 (local) | $0.002 | $0.003 |
| Built-in Q&A | Yes | Yes | No | No |
| Local Deployment | Advanced | Easy | Possible | No |
| License | Apache 2.0 | Apache 2.0 | MIT | Proprietary |
Ready to Use Voxtral in Your Projects?
Get started with Voxtral's open-source models and build the next generation of speech AI applications.