Playground For LM

Run local MLX models, understand images, use RAG, and route smarter workflows with Playground for LM — now with beta APIs for advanced AI setups.

Your chats, voice transcriptions, and files stay on your device.

Playground for LM is a private local AI workspace for Apple devices. Run supported language models with local llama.cpp-based inference, chat with models offline, manage your own model library, and transcribe supported audio locally with Whisper.cpp.

The app includes a limited free trial so you can test the core workflow on your device. Playground For LM Pro unlocks the full experience, including extended local model usage, full transcription access, premium themes, and the exclusive gilded Pro app icon.

Local AI Chat

Run supported local language models directly on your device. Test models, compare responses, experiment with prompts, draft text, and build personal AI workflows without relying entirely on cloud-based chat services. Free users can try the chat workflow with a limited message allowance. Playground For LM Pro removes the free-user message restriction.

Local Audio Transcription

Turn supported recorded or imported audio into text with local Whisper.cpp-powered transcription. Review generated transcripts and organize spoken content into cleaner, more readable text. Free users can try transcription with a limited allowance. Playground For LM Pro unlocks the full transcription workflow.

Model Library Management

Choose or create your own model library folder. Add models, load them, chat with them, and unload them when you are done. Playground for LM is designed for users who want clear control over where their models are stored and how they are used.

Clean Output Rendering

Playground for LM supports readable rendering for common AI-generated formats, including Markdown, code blocks, LaTeX formulas, tables, and structured lists.

System Diagnostics

View useful runtime and model information while working with local AI. Depending on your device, model, and workflow, the app may display loading state, model status, memory-related notes, transcription state, runtime behavior, and generation performance.

Playground For LM Pro

Upgrade to Playground For LM Pro to unlock extended local model usage, full local transcription access, premium themes, the exclusive gilded Pro app icon, and future Pro features.

Privacy-Oriented by Design

When using supported local language models, inference runs on your device through the app’s local runtime. When using local transcription, supported audio is processed on device.

Some actions, such as downloading models, importing files from external locations, restoring purchases, completing in-app purchases, or opening external legal documents, may require a network connection.

Important Notes

Playground for LM does not include every possible AI model or transcription model by default. Users are responsible for ensuring that any model they use is compatible with the app and that they have the right to use it.

Local AI models and transcription models may require significant storage, memory, and processing power. Larger models may load slowly, fail to load, or perform differently depending on your device.

Generated content and transcription results may be inaccurate, incomplete, or misleading. Speech transcription may be affected by background noise, speaker accent, audio quality, language, and model limitations. Always review important information carefully before relying on it.

Playground for LM gives you a practical way to experiment with private local AI on Apple devices: local chat, Whisper.cpp-powered transcription, model management, clean rendering, diagnostics, and a focused workspace under your control.