stev-dl·3 years ago·discussLooks like they're adding in quite a few "multimodal" features in GPT-5. Emphasis on audio: artificial speech production, audio-to-text, voice recognition - likely building on Whisper. Translation for text/speech also seems on the roadmap.