πŸ‡©πŸ‡ͺ New German STT & Improved PII Detection Models

German is now supported in Universal-1, our top multilingual STT model! Transcribe German audio and video with high accuracy. We've also enhanced PII Redaction and Entity Detection, adding 16 new entities and support for Chinese, Dutch, Japanese, and Georgian.

πŸ‡©πŸ‡ͺ New German STT & Improved PII Detection Models

Hey πŸ‘‹, this weekly update contains the latest info on our new product features, tutorials, and our community.

πŸ‡©πŸ‡ͺ German Speech-to-Text Added To Universal-1

German language support is now available in Universal-1, our most powerful and accurate multilingual Speech-to-Text (STT) model! This launch makes it easier to transcribe German audio and video with high-quality speech recognition.

Get Started:

To help you integrate Universal-1 into your applications, check out these docs and guides: 

Join the Conversation

We’re excited to see what you build with Universal-1’s enhanced German support. Share your projects and join the conversation on Twitter with the hashtag #AssemblyAI and mention us @AssemblyAI.

Improved PII and Entity Detection models

We've improved our English model for PII Text Redaction and Entity Detection, increasing accuracy in detecting and removing personally identifiable information and other entities from transcripts.

New features include:

  • 16 New Entities: Including vehicle_id and account_number
  • Language Support: Added Chinese, Dutch, Japanese, and Georgian for PII Text Redaction, now covering a total of 47 languages. 

Generate transcripts with sensitive information removed using our audio intelligence models. For more details, check out our documentation.

Fresh From Our Blog

Redact Personal Identifiable Information (PII) from audio with Node.js: Automatically redact Personal Identifiable Information from audio and video files with Node.js and the AssemblyAI API. Read more>>

Extract phone call insights with LLMs in Python: Learn how to automatically extract insights from customer calls with Large Language Models (LLMs) and Python. Read more>>

Summarize audio with LLMs in Node.js: Summarize audio with Node.js and the AssemblyAI speech-to-text API. Read more>>

Best way to transcribe and sift through large amounts of audio: Human transcribers cost somewhere between 15 to 40 dollars per hour of audio and take hours to transcribe. With AssemblyAI's Universal-1 model, you can now transcribe at a fraction of the cost and at lightning speed.

Coding an AI Voice Bot from Scratch: Real-Time Conversation with Python: Build a real-time AI voice assistant using Python that can handle incoming calls, transcribe speech, generate responses, and provide a human-like conversational experience.

How to use @postman to test LLMs with audio data (Transcribe and Understand): Learn how to transcribe audio and video files using AssemblyAI and also how to use LeMUR, AssemblyAI's framework for using Large Language Models on spoken data without having to code at all.