OpenAI just released new, more accurate audio models to replace Whisper:

Whisper often struggles with transcriptions in LingQ, especially for imports with accents, background noise, or certain languages. These new models are faster, more multilingual, and more accurate.

Could LingQ consider switching to them? It would improve import quality and give users a smoother experience when working with audio. This seems like a great upgrade to keep LingQ ahead of the curve.

Would love to hear your thoughts—thanks!