1 article tagged with this topic
ll ama-server now supports speech-to-text via Google's Gemma-4 E2A and E4A multimodal models.