AssemblyAI

AssemblyAI is a voice AI infrastructure company that sells speech to text and speech understanding through an API. Developers send prerecorded audio or a live stream and receive transcripts, and can layer on structured understanding such as entity detection, topic detection and translation, guardrails such as PII redaction, profanity filtering and content moderation, and a full Voice Agent API for building speech to speech agents over WebSocket. An LLM Gateway routes requests to third party language models under the same account. It is built for engineering teams that need accurate transcription at production scale: meeting and call recording products, live captioning, contact center agent assist, media indexing and voice agents. The product lineup is a Prerecorded Speech to Text API, a Real time Speech to Text API, the Voice Agent API, a Speech Understanding API, Guardrails and the LLM Gateway, backed by the Universal family of models. HIPAA business associate agreements, PCI DSS and SOC 2 compliance are included without a premium. Pricing is published per hour of audio and add ons stack. Prerecorded transcription costs $0.21 per hour on Universal 3.5 Pro and $0.15 per hour on Universal 2. Streaming costs $0.45 per hour base on Universal 3.5 Pro Realtime and $0.15 per hour on Universal Streaming for English and multilingual audio. The Voice Agent API is $4.50 per hour, or $0.075 per minute, inclusive of speech to text, the language model, text to speech and orchestration. Add ons include Medical Mode at $0.15 per hour, speaker diarization at $0.02 per hour for standard batch jobs and $0.12 for streaming, Voice Focus at $0.10, entity detection at $0.08, topic detection at $0.15, translation at $0.06, PII text redaction at $0.08, content moderation at $0.15 and profanity filtering at $0.01. The LLM Gateway is billed per million tokens by model. New accounts receive $50 in free credits with no credit card required.