AssemblyAI

AssemblyAI is a speech to text API rather than an app: you send audio and get back a transcript, with speaker labels and summarisation available on the same call. Both pre-recorded and live streaming are covered, and the free tier is generous enough to build something real before paying anything. It is bought by people writing software, not by people who want a transcript this afternoon.

Usage priced with no monthly plan at all. Universal-2 is $0.15 an hour and Universal-3.5 Pro, the more accurate model, is $0.21 an hour, with streaming models priced on their own line. The free tier covers 185 hours of pre-recorded transcription and 333 hours of streaming, which is far past what most projects need to prove themselves. Streaming concurrency starts at 100 sessions a minute on pay as you go and scales automatically as you approach the limit, at no extra cost.

Somebody writing software that needs transcripts, whether that is a tool for creators or an internal pipeline. If you want a transcript of an interview this afternoon and have no intention of writing code, every other listing in this category is a better answer.