Run powerful language and mixture-of-experts models entirely on your Mac. Use MLX, stream large MoE experts from SSD, or use Apple Intelligence — all from one native chat app.
Download the app. Pick a model. Chat. No inference servers. No clients. Just local AI that feels like an app.
Qwen, Gemma, Llama, Phi, DeepSeek — quantized for Apple Silicon. Import any MLX-compatible model from Hugging Face.
Zero downloads. Zero setup. Future Apple Intelligence upgrades come to MLX Chat automatically.
Watch reasoning models think in real time, with collapsible thinking blocks and live throughput metrics.
The model writes HTML or SVG; it renders beside the chat in a sandboxed web view. Markdown, syntax highlighting, and LaTeX included.
Download with progress and cancel. Delete cleanly. MLX stays the default; Gemma 4 26B and Qwen3.6 35B can keep their shared core in memory and stream only routed experts from a resumable .gturbo install on SSD.
Idle models automatically give memory back after 2 minutes, 10 minutes, an hour, or never. When they're unloaded, the memory actually comes back.
Quantized models load straight into Apple Silicon's unified memory and stay hot between turns.
ChatSession reused across turns — no re-prefilling the conversation Gradient descent is how a model learns: feel the slope under your feet on a foggy hillside, step downhill, reassess. Each step updates the parameters θ against the gradient of the loss L:
var theta = initialGuess for _ in 0..<steps { theta -= learningRate * lossGradient(at: theta) }
Keep the shared core and FP16 KV cache in memory, then read only router-selected experts from SSD.
The FoundationModels framework hands MLX Chat the system's on-device model. Nothing to ship, nothing to download, nothing to babysit.
LanguageModelSession, fresh per turn Download the DMG or build from source. On Apple Intelligence–enabled Macs, chat works the moment the app opens — no model download needed.
Drag to Applications. Chat with Apple Intelligence out of the box, or pull a local model from the built-in catalog.
Download DMGRequires an Apple Silicon Mac, macOS 26 with Xcode 26, and XcodeGen.
# clone and generate git clone https://github.com/abayomipopoola/mlx-chat cd mlx-chat && xcodegen generate # build (mlx-swift ships a plugin + macros) xcodebuild -scheme MLXChat -configuration Release \ -skipPackagePluginValidation \ -skipMacroValidation build