OpenAI looks to be aggressively deprecating old models, including models for which there is no adequate replacement.
For example, they're deprecating all their TTS models with no suitable replacement. It would be completely absurd to have to use the recommended gpt-realtime-2.1-mini model for TTS instead. It is obvious they have no idea what they're doing. Their documentation pages are chock-full of inconsistent information. Moreover, going forward, theirs looks to be an ecosystem that one cannot place trust in.
OpenAI looks to be aggressively deprecating old models, including models for which there is no adequate replacement.
For example, they're deprecating all their TTS models with no suitable replacement. It would be completely absurd to have to use the recommended gpt-realtime-2.1-mini model for TTS instead. It is obvious they have no idea what they're doing. Their documentation pages are chock-full of inconsistent information. Moreover, going forward, theirs looks to be an ecosystem that one cannot place trust in.