LM Studio Launches Local Real-time Voice Transcription
LM Studio introduces a local real-time voice transcription feature, allowing users to interact with agents without cloud data transmission, now supported on Mac
On September 16, 2026, LM Studio officially introduced a local real-time voice transcription feature. This update enables users to communicate freely with their agents without the need to transmit sensitive voice data to the cloud.
Image source: @lmstudio
This new capability focuses on protecting user privacy, ensuring all processing happens directly on the user's device. Beyond the existing support for Mac and Windows, this release expands compatibility to Linux, allowing a broader range of users to utilize real-time voice interfaces within their local AI environments.
Privacy-Focused Local Voice Transcription
LM Studio’s new transcription feature utilizes a 'local on-device' processing method. By completing all transcriptions locally, users can interact with AI agents using voice without the privacy concerns typically associated with cloud-based voice data transmission.
Expanded Support for Linux
Alongside the established support for Mac and Windows platforms, this update extends the real-time voice transcription capability to Linux environments. This improvement enhances accessibility and provides a more unified local AI experience for developers and users operating across different operating systems.