customers
All customer stories
Top Voice AI companies are building with Assembly.
resources
Latest Release
Dictation API
The first API built for dictation. Your users speak, and it returns text that is ready to send: filler gone, self-corrections resolved, names spelled right.
resources
Multilingual speech-to-text with speaker diarization, real-time transcription, and LLM integrations for intelligence out of the box—best-in-class word error rate from $0.15/hr.
Welcome back to the show—today we're joined by the team behind Northwind.
Thanks for having me. We process about 2 million minutes of audio a month.
That's a lot. Let's get into how you keep accuracy high at that scale.
Best-in-class word error rate on our published benchmarks: Universal-3.5 Pro for 18 languages with natural language prompting, Universal-2 for 99 languages at $0.15/hr.
Identify and label speakers on every audio file, streaming or pre-recorded.
Pipe clean, formatted transcripts straight into LLMs for summarization, topic extraction, and downstream workflows.
36%
improvement in close rate
“The new Universal-3.5 Pro speech model from AssemblyAI is best so far in terms of accuracy, latency, and language switching.”
80%
increase in customer satisfaction
“Assembly has saved us countless hours managing models, and provided exceptional accuracy.”
36%
improvement in close rate
“The new Universal-3.5 Pro speech model from AssemblyAI is best so far in terms of accuracy, latency, and language switching.”
80%
increase in customer satisfaction
“Assembly has saved us countless hours managing models, and provided exceptional accuracy.”
Create a free account and transcribe any file with one API call. Test it in the no-code playground, then copy a ready-made request into your app. Usage-based pricing from $0.15/hr for Universal-2 and $0.21/hr for Universal-3.5 Pro, no minimums.
Start building freeNo credit card required
Word error rate is the share of words the model gets wrong against a human reference—the standard measure of Speech-to-Text accuracy on pre-recorded audio.
Average normalized WER across synthetic medical, accented English, general speech and webinar sets.
*Lower is better*Source: AssemblyAI published benchmarks (assemblyai.com/benchmarks), pre-recorded tab, updated July 2026.
