Minimalist AI Pro Support
Minimalist AI Pro runs AI models on your iPhone, iPad or Mac. Chat, image generation, transcription and speech stay on your device.
Getting started
- Chat: open Settings › Models & Files, then tap Download Recommended Model or import a compatible GGUF file, MLX folder or MiniCPM package. Load it and start chatting. The thinking button, image input and JS/Py tools follow what the loaded model supports.
- Images: in the Image workspace, open Settings › Download Recommended Model. Download Stable Diffusion 1.4 or 1.5 (Apple Core ML) or Bonsai Image 4B after accepting the model license. You can also import Apple Core ML Stable Diffusion 1.x / 2.x model folders (512×512) there.
- Audio: in the Audio workspace, choose an audio or video file to transcribe on device, or type text and use Read Aloud, or Generate Audio to create a WAV file you can listen to and save. Transcription needs iOS, iPadOS or macOS 26 or later; Apple's recognition resources for a language can be downloaded in Settings.
- Make it yours: Settings › Personalization sets the workspace background, Soften Background, Component Transparency and bubble colors. Settings › Names lets the assistant use your name and its own; Advanced parameters › System Prompt adds your own instructions. Use the magnifier in the chat list to search chats.
Requirements
- Image generation needs a 12 GB-class iPhone or iPad, or a Mac with 16 GB of memory or more.
- Larger chat models need more memory; each recommendation lists its size and memory class.
- Models are not included in the app download.
Common questions
A model will not load. Check free storage and memory, close other large apps and try a smaller model. Imported files must be a supported format and complete.
Downloads stop. Downloads can be paused and resumed. Settings › Downloads › Allow Cellular and Metered Connections controls whether model downloads may use cellular data (on by default).
The JS or Py Tool did not run. Turn the tool on in Settings and in the chat controls. The model decides when to call it; some models do not support tool calls.
An image was not saved. Every image is checked by a safety checker; flagged results are discarded.
Privacy when reporting a problem
Do not send private chats, confidential documents, model files, passwords or API keys. Reproduction steps and an error screenshot are usually enough. See the Minimalist AI Pro Privacy Policy.
Contact
Email LuoHanjunElvis@gmail.com, or use Settings › Contact the Developer in the app.
Minimalist AI Pro 支持
Minimalist AI Pro 在你的 iPhone、iPad 或 Mac 上运行 AI 模型，聊天、图片生成、转写和朗读都在本机完成。
开始使用
- 聊天：打开设置 › 模型与文件，点“下载推荐模型”，或导入兼容的 GGUF 文件、MLX 文件夹或 MiniCPM 模型包。加载后即可聊天。思考按钮、图片输入和 JS/Py 工具会按所加载模型实际支持的能力显示。
- 图片：在图片工作区打开设置 › 下载推荐模型，同意模型许可后即可下载 Stable Diffusion 1.4 或 1.5（Apple Core ML）或 Bonsai Image 4B。也可以在这里导入 Apple Core ML 格式的 Stable Diffusion 1.x / 2.x 模型文件夹（512×512）。
- 声音：在声音工作区选择音频或视频文件进行本地转写，或输入文字后使用“直接朗读”，或“生成音频”得到可试听和保存的 WAV 文件。转写需要 iOS、iPadOS 或 macOS 26 及以上版本；各语言的 Apple 识别资源可以在设置中下载。
- 个性化：设置 › 个性化可调整工作区背景、柔化、组件透明度和聊天气泡颜色；设置 › 名称可让助手称呼你并使用它自己的名字；在系统提示词中可以加入你自己的要求。在聊天列表中点放大镜可以搜索会话。
设备要求
- 图片生成需要 12 GB 档的 iPhone 或 iPad，或 16 GB 及以上内存的 Mac。
- 较大的聊天模型需要更多内存，每个推荐模型都标有大小和内存档位。
- App 下载不包含模型。
常见问题
模型无法加载。请检查剩余存储和内存，关闭其他大型 App，或换用较小的模型。导入的文件必须是受支持的格式且完整。
下载中断。下载可以暂停和继续。设置 › 下载 › “允许蜂窝及按流量计费的连接”决定模型下载是否可以使用蜂窝网络（默认开启）。
JS 或 Py 工具没有运行。请在设置和聊天控件中开启工具。是否调用由模型决定，部分模型不支持工具调用。
图片没有保存。每张图片都会经过安全检查，被标记的结果会被丢弃。
报告问题时的隐私
请不要发送私人会话、机密文件、模型文件、密码或 API 密钥。复现步骤和错误截图通常就足够了。参见 Minimalist AI Pro 隐私政策。
联系
发邮件至 LuoHanjunElvis@gmail.com，或在 App 中使用设置 › 联系开发者。
Minimalist AI Pro