AssemblyAI

AssemblyAI provides speech recognition through an API that developers can plug into their own products. Beyond plain transcription, it can label speakers, detect topics, summarize recordings and flag sensitive content. It supports both pre-recorded files and real-time streaming, and offers a way to ask language-model questions about a transcript. Its models are trained for accuracy across accents and noisy audio. Pricing is pay as you go per hour of audio, with free credits to start. It suits companies building call analytics, meeting tools and media products.