OpenAI's hosted Whisper-3 endpoint is the easy default - same model architecture as WhisperX uses, just hosted with the OpenAI API ergonomics you already know.
No word-level alignment out of the box (you'd add WhisperX or verbose_json post-processing); good language coverage; ~$0.006/minute pricing.
For teams without GPU infra, this is the convenient hosted version of the open-source champion.
Verdict: convenient hosted Whisper. 8.4/10.