OpenAI to shut down legacy TTS models tts-1, tts-hd and gpt-4o-mini-tts on January 6, 2027

OpenAI is deprecating its legacy text-to-speech models tts-1, tts-hd and gpt-4o-mini-tts, with access ending January 6, 2027. Developers who built services on those voices, including the Listen Later app, must rebuild around alternatives such as the ElevenLabs v4 engine.

• Access to tts-1, tts-hd and gpt-4o-mini-tts ends January 6, 2027
• Listen Later, built entirely on these voices, loses its core product
• Six Colors Audio Newsletter is switching to the ElevenLabs v4 engine
• A new workflow re-transcribes audio to catch omissions and hallucinations