customers
All customer stories
Top Voice AI companies are building with Assembly.
resources
Latest Release
Universal 3.5 Pro
Our new flagship speech-to-text model handles real-world audio the way it actually happens. Available for real-time and pre-recorded audio.
resources
Medical Mode is a domain add-on for our speech-to-text API that catches the drug names, dosages, and conditions general models miss—reducing medical entity errors by 87% versus the base Universal-3 Pro model. Set one parameter, no pipeline changes.
Let's keep you on Lisinopril 10 milligrams daily for the blood pressure.
Got it. And the one for my blood sugar?
Metformin, twice a day with meals.
Tuned for medical entities, it corrects medication names, procedures, conditions, and dosages that general models get wrong.
Turn it on by setting the domain to medical-v1 on the same async or streaming call—no changes to your existing pipeline.
Pairs with speaker diarization and PII redaction. A Business Associate Addendum (BAA) is available and signable in minutes without a sales call—plus SOC 2 Type 2 and ISO 27001:2022.
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.”
Add one parameter—set the domain to medical-v1—to any pre-recorded or streaming request and Medical Mode applies automatically. Sign up free, drop in your API key, and pipe clean transcripts into an LLM for SOAP notes or summaries.
Start building freeNo credit card required
On medical audio, missed entity rate measures how often medications, dosages, and conditions are dropped or garbled—the errors that break a downstream note.
Missed entity rate on medical audio.
*Lower is better*Source: AssemblyAI published benchmarks — assemblyai.com/benchmarks.
