Translate design frames, rephrase UI copy, and access AI writing assistance directly within Figma to streamline your localization and content creation workflow.
Ethertext
Transform your clipboard text instantly using advanced AI models for grammar fixes, code debugging, and style changes directly within your macOS workflow.
Free

About Ethertext
Ethertext is a macOS utility designed to integrate generative AI directly into the system clipboard. It allows users to copy text, apply a variety of AI-driven transformations, and then paste the modified result. The tool acts as a bridge between standard productivity apps and powerful Large Language Models (LLMs), eliminating the need to manually copy-paste content into browser-based chat interfaces. By living in the menu bar, it provides a seamless "Copy, Transform, Paste" workflow that works across any application on the Mac. The tool's functionality extends beyond simple text rewriting. Key features include one-shot text cleaning via keyboard shortcuts, voice dictation with automated transcription, and a "Memorize" function that stores text or webpage content for later retrieval. A unique "Recall" feature allows users to query their past copied or memorized data using AI assistance. For developers, Ethertext offers specialized prompts for explaining, debugging, and translating code snippets. The application supports a wide range of models, including the latest from OpenAI, Google, and Anthropic, as well as local execution via Ollama and Apple's MLX framework for enhanced privacy. Ethertext is particularly well-suited for writers, developers, and researchers who frequently move text between applications. Professionals can use it to instantly shift the tone of an email, while programmers can use it to refactor code snippets without leaving their IDE. Because it supports local models through MLX and Ollama, it also appeals to privacy-conscious users who prefer to process sensitive data on-device rather than sending it to third-party cloud providers. What distinguishes Ethertext from typical AI writing assistants is its flexibility and technical depth. Instead of being locked into a single provider, users can bring their own API keys or run models locally on Apple Silicon. The inclusion of system-wide keyboard shortcuts and specialized capture tools—such as multi-monitor screenshots with OCR—makes it a deeply integrated power tool rather than a standalone wrapper. Its recent updates focus on "on-device" capabilities, leveraging the MLX framework to ensure that embeddings and chat models can run entirely offline.
Ethertext pros & cons
Pros
- Supports local on-device AI models via MLX for maximum privacy and no per-request API costs.
- Seamlessly integrates with any macOS application via system-wide keyboard shortcuts.
- Offers broad compatibility with multiple major AI providers including OpenAI, Google, and Anthropic.
- Includes vision and audio capabilities for transcribing screenshots and voice memos directly into the clipboard history.
- Allows for high customization with user-defined prompts and adjustable model temperature settings.
Cons
- Only available for macOS users, with no support for Windows, Linux, or mobile platforms.
- Requires users to provide their own API keys for cloud-based AI functionality.
- Setting up local models via Ollama may require technical familiarity with command-line tools like Homebrew.
- Limited to macOS 14 and newer, which may exclude users with older hardware.
Ethertext use cases
- Software developers can highlight code and use shortcuts to instantly debug, explain, or refactor snippets without leaving their editor.
- Content creators can quickly shift the tone of drafts to different styles, such as 'Gen Z' or 'Formal', across any writing app.
- Researchers can use the 'Memorize' and 'Recall' features to capture information from multiple sources and query it later with AI assistance.
- Busy professionals can use the dictation shortcut to record and transcribe thoughts into text that is automatically stored for future recall.
- Writers can use the 'Clean Text' shortcut (⌘⇧T) to instantly fix grammar and spelling in any text field on their Mac.
Ethertext features
- custom prompt creation
- voice dictation and transcription
- system-wide keyboard shortcuts
- support for gpt-5.2, gemini 3, and claude 4.5
- ai-powered memorization and recall
- screen capture with ocr
- local ai model support (mlx & ollama)
- one-click text transformations
Ethertext pricing
Is Ethertext free? Yes, Ethertext has a free plan.
Free
Free
- Local AI support (MLX & Ollama)
- OpenAI, Gemini, & Anthropic integration
- System-wide keyboard shortcuts
- Voice dictation and transcription
- Screenshot OCR and vision
- Text memorization and recall
- Customizable AI prompts
- Adjustable creativity/temperature settings
Ethertext FAQs
How do I use my own AI models with Ethertext?
You can enter your own API keys for OpenAI, Google Gemini, or Anthropic in the application settings to use cloud-based models. Alternatively, you can enable Ollama support or use built-in MLX models for local processing.
Can Ethertext run without an internet connection?
Yes, Ethertext supports local AI via the MLX framework on Apple Silicon and Ollama integration. This allows you to perform text transformations and embeddings entirely on your device for improved privacy.
What is the Recall feature and how does it work?
Recall allows you to query your previously memorized text, webpages, and screenshots using AI. It uses embedding models from providers like OpenAI or local Ollama instances to find and retrieve relevant information from your history.
Does Ethertext support image-to-text transcription?
Yes, using the ⌘⇧S shortcut, you can capture a screenshot which Ethertext will then transcribe and memorize using Apple's local Vision model.
What are the system requirements for Ethertext?
Ethertext is designed specifically for macOS and requires version 14 (Sonoma) or newer to run. It is optimized for Apple Silicon to take advantage of on-device MLX models.
Ratings & reviews
No reviews yet. Be the first to share how Ethertext worked for you.