ment

empowering your ai experience with privacy and control

Discover why ment is the ideal choice for private, on-device AI:

local large language model (lllm)

ment runs entirely on your Mac, delivering:

• complete privacy: no conversations ever leave your device
• enhanced security: data stays local so there are no third‑party servers to compromise
• reduced latency: responses are generated instantly without network delays

secure conversation storage

Every conversation and assistant profile is encrypted with AES‑GCM using a key stored in your macOS/iOS Keychain.

model manager

Choose the model that suits your needs:

• multiple models: add or remove HuggingFace/MLX models at any time
• offline availability: once downloaded, models work without internet access

conversation tools

Stay in control of the dialogue:

• branching threads: explore alternate replies and switch between them
• message editing: revise previous prompts and regenerate answers
• generation stats: view token counts, model used and speed for each response

customization & flexibility

Create assistants tailored to your workflow:

• system prompts: specify a role and tone for each profile
• favorites: mark go‑to models and profiles for quick access
• rich formatting: support for code blocks, lists, quotes and links in replies

upcoming feature: pdf integration

Soon you’ll be able to import PDF documents so the AI can reference them directly within your conversations.

ment is continually evolving—expect more features that keep your AI powerful, private and fully in your hands.