AssemblyAI
1 App · 7 Intel
AssemblyAI is a voice AI company that builds speech recognition and speech understanding models and offers them to developers and enterprises through APIs. Dylan Fox founded it in 2017, and it went through Y Combinator that same year. The company has offices in New York and San Francisco, and its funding includes a $50 million Series C led by Accel in 2023.
Is this you? Sign in with X to claim this profile.
Apps
Open-source Mac dictation app that pastes cleaned-up speech into any focused app via AssemblyAI's Dictation API.
Intel
AssemblyAI released Vet Mode in beta, a post-transcription correction pass enabled with domain=veterinary-v1. It fixes vet drug names and anatomy terms that general models swap for human drugs, supports streaming and pre-recorded audio, and costs $0.15/hr extra.
AssemblyAI's Universal-3.6 Pro transcription model is now available on OpenRouter through its Sync API. One request returns a finished transcript without polling, webhooks or WebSockets, so existing OpenRouter users can add speech-to-text on the same key. Pricing and limits were not in the source.
Want an AI voice agent that joins your Zoom, Meet or Teams call and answers questions out loud? @recallai just published the full build. AssemblyAI's Voice Agent API listens, thinks and speaks. https://t.co/az8yXeqYRF's Output Media puts it in the meeting. https://t.co/eLSDBVaF1B What's inside: - One WebSocket for Speech-to-Text, LLM and voice - Short-lived tokens, so your API key stays out of the browser - Far-field Voice Focus for laptop and room mics - Tool calls, up to leaving the meetin
AssemblyAI launched a Dictation API on Universal 3.5 Pro. A demo fills a patient form from one spoken answer, flags unanswered fields, and supports key-term lists and prompts for output style such as email or bullets.
A walkthrough of AssemblyAI's Dictation API playground shows how the key-terms field locks in domain vocabulary (drug names, SKUs, product names) and reports sub-second turnaround on Universal-3.5 Pro across 19 languages.
AssemblyAI releases Universal-3.6 Pro Realtime, a streaming speech-to-text model for voice agents. It offers 32 languages from one endpoint, entity-aware endpointing, and background-voice suppression, and AssemblyAI says it leads the Pipecat and Coval streaming benchmarks.
AssemblyAI launched a Dictation API on Universal-3.5 Pro. One request returns cleaned text with fillers removed and custom formatting instructions applied, across 19 languages, sub-second on short clips, at $0.62 per hour. Blurt is the reference push-to-talk app.