OpenAI has introduced two new transcription models to its API, according to its official announcement. The first model, GPT-Live-Transcribe, is built for low-latency live transcription. The second, GPT-Transcribe, is optimized for asynchronous transcription of completed audio files and batch workloads.
Key Facts
- GPT-Live-Transcribe: designed for live transcription with low latency
- GPT-Transcribe: optimized for asynchronous transcription and batch processing
- Both models are now available via the OpenAI API
- Announcement made via OpenAI's official Bluesky account
What It Means
The split into two specialized models could be relevant for German enterprises that need real-time transcription (e.g., in customer service or meetings) or large-scale audio processing. Pricing differences and exact performance metrics remain to be clarified.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
