Dictation API by AssemblyAI
Product Hunt Devon Malloy
AssemblyAI launched a Dictation API that turns speech into polished text with one API call. It strips filler and false starts, and can format the result as notes, replies, or commit messages.
Based on reporting by Product Hunt, Devon Malloy — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
AssemblyAI is adding a Dictation API to its speech stack, and the pitch is simple: make spoken clips come out as finished text with almost no fuss. A single API call is supposed to do the job.
The output is not just a raw transcript. AssemblyAI says the system removes filler words and false starts, then shapes the text the way the user wants it — whether that means notes, a commit message, or a customer reply. That’s a useful distinction, because a lot of dictation tools still leave the cleanup work to the human.
The service is built on Universal-3.5 Pro and supports 19 languages. On short clips, AssemblyAI says it can return results in under a second, with pricing set at $0.62 per hour all in.
This is the company’s 11th launch, which fits the broader picture here: AssemblyAI keeps pushing deeper into voice AI rather than just selling transcription. The same product family already includes speaker detection, summarization, PII redaction, an LLM gateway, and a Voice Agent API, plus async and real-time streaming support for notetakers, medical scribes, call analytics tools, and other voice apps.
My take — AI-written commentary, not fact-checked reporting
This is the right kind of boring: less “AI magic,” more text that’s ready to use. The market has plenty of models that can hear you; the real test is whether they can stop making humans do the cleanup afterward. AssemblyAI seems to understand that the dull part is the product.
Read more about this at: Product Hunt