Google Launches Open-Source Voice AI Stack Powered by Gemma 4 31B Model
Google, Hugging Face and Cerebras have launched an open-source speech-to-speech stack that uses the 31B parameter Gemma 4 model for voice artificial intelligence. The combined system delivers what the companies describe as ultra-fast inference speeds for developers building conversational applications.
The framework integrates a fully open-source cascade that can be paired with existing voice platforms. A working demo and technical documentation are now available for developers to test the integration.
From the sources (5 posts)
@googlegemmaVoice AI without the wait! ⏱️ Thanks to Hugging Face and Cerebras, developers can now use the Gemma 4 31B model as the brain for voice AI at ultra-fast inference speeds. Add it to a fully open-source, cascaded speech-to-speech stack that c
@googlegemmaTry the demo by @andimarafioti and team: Read the blog:
@huggingfaceRT @googlegemma: Voice AI without the wait! ⏱️ Thanks to Hugging Face and Cerebras, developers can now use the Gemma 4 31B model as the br…
@victormustarRT @googlegemma: Voice AI without the wait! ⏱️ Thanks to Hugging Face and Cerebras, developers can now use the Gemma 4 31B model as the br…
@andimarafiotiRT @googlegemma: Voice AI without the wait! ⏱️ Thanks to Hugging Face and Cerebras, developers can now use the Gemma 4 31B model as the br…