App uses OpenAI tts-1? Text-to-speech models shut down January 6, 2027
Ship ItChecked against the official source
Source checked 28 Oct 2026 · Published 29 Oct 2026 · Last reviewed 29 Oct 2026 · By Ctrl+Z Therapy (how we test)
The short answer
OpenAI's deprecations page says these text-to-speech models are deprecated and will be removed from the API on January 6, 2027: tts-1, tts-1-hd, and both gpt-4o-mini-tts snapshots (gpt-4o-mini-tts-2025-03-20 and gpt-4o-mini-tts-2025-12-15). OpenAI announced it on October 1, 2026, with at least three months' notice. The recommended replacement for each is gpt-realtime-2.1-mini, and OpenAI says to migrate before the shutdown date.
What to do
- Find every place your app calls OpenAI's text-to-speech and check which model it uses.
- If it's tts-1, tts-1-hd or gpt-4o-mini-tts, switch the model to gpt-realtime-2.1-mini before January 6, 2027.
- Ask your AI tool: "Find every OpenAI text-to-speech call in this app. If it uses tts-1, tts-1-hd or gpt-4o-mini-tts, show me what changes to move to gpt-realtime-2.1-mini, then test one."
Good to know
- Our tip: listen to the result before you ship. A different model can sound different and may need different settings.
Source
Checked on 28 Oct 2026. Prices, limits and settings change, so check the source before you rely on them.