LM Studio Bionic

LM Studio Bionic is an agent for work and code designed to run locally on a user’s device, enabling knowledge workers and developers to create, edit, and automate tasks. It emphasizes local processing, privacy, and support for frontier models, making it suitable for individuals who need an on-device assistant and coding helper. The tool integrates with open models and offers a path to use frontier AI while keeping control of data.

The system operates by providing an agent that can create documents, code, and automated tasks, with real-time voice transcription and local processing. It highlights the ability to work with documents, coding tasks, and automations, while ensuring voice data remains on device. The design centers on enabling creativity, work, and code with an on-device approach.

LM Studio Bionic supports voices, transcription, and multiple languages, and it allows users to run local LLMs directly within the app. It uses the LM Studio runtime with MLX and llama.cpp, and it can connect to frontier models for demanding tasks. The architecture emphasizes local operation and on-device control of models.

The product outlines plans for local and cloud usage, with a cloud option labeled as Zero Data Retention (ZDR). It notes frontier model support, including GLM 5.2, Kimi K3, and DeepSeek V4 Pro, and mentions an enterprise path for teams with options to deploy on own infrastructure or edge devices. It frames privacy as an essential aspect of its ethos.

Use cases

• Automate document creation and editing, coding tasks, and simulations.
• Voice-enabled workflows with real-time transcription for multilingual use.