Search

AssemblyAI

Freemium

AssemblyAI sells speech models through an API rather than a consumer app. Developers send a file or live stream and get transcripts with timestamps, speaker labels, formatting and confidence scores; other endpoints summarize calls, detect topics and sentiment, or redact personal data. Its Universal speech models cover 99 languages, and a newer text-to-speech model targets voice agents. Billing runs per hour of audio processed, starting at roughly $0.15 per hour for pre-recorded transcription. The free tier covers about 185 hours of pre-recorded audio and 333 hours of streaming before charges begin.

Murf AI

Free trial

Murf combines a browser voiceover editor with a speech API. In the studio you paste a script, assign a voice to each paragraph, adjust pitch, pace and emphasis, then export audio for video or e-learning. It offers over 200 voices across 30+ languages, plus AI dubbing for existing footage. The API exposes text-to-speech billed per 1,000 characters, speech-to-text at $0.24 an hour, voice changing and translation, with $10 of trial credit. Studio plans start at Creator; Business adds 96 hours of generation, 48 hours of transcription and a business license for teams.