OpenAI releases new speech recognition models

OpenAI has added two new speech recognition models to its API. These tools are designed to convert spoken language into text more efficiently than before.

While this is a step up from their older technology, the performance still trails behind some competitors. Benchmarks show that models from ElevenLabs, Google, and Mistral are currently more accurate.

If you use these tools for building apps, you might find better error rates elsewhere for now. OpenAI will likely keep refining these models to close the gap.