Emacs org-mode package
FreeEmacs as your personal AI assistant
FreeFree tier
Inputs: textOutputs: text, image
About Emacs org-mode package
org-ai is an Emacs org-mode minor mode that integrates generative AI models directly into the editing environment. It supports OpenAI's ChatGPT and DALL·E, Stable Diffusion via stable-diffusion-webui, and offers speech input/output. Users can generate text, images, and image variations using org-mode special blocks, control system and user prompts, and leverage commands outside org-mode for selected text or multiple files. The package requires an OpenAI API key or similar backend setup and is installable via Melpa or straight.el.
Key Features
Generative AI inside org-mode buffers with #+begin_ai...#+end_ai blocks
Supports ChatGPT, DALL-E, and other OpenAI text models
Image generation and variation via DALL-E and Stable Diffusion
Speech input and output for talking with the AI
Commands usable outside org-mode for selected text or multiple files
Noweb support for embedding AI blocks in literate programming
Customizable system and user prompts per block
Syntax highlighting in AI blocks and auto-fill on insertion
Backend flexibility: OpenAI, Azure, Anthropic, perplexity.ai, and local LLMs via oobabooga/text-generation-webui
Pros & Cons
Pros
- Deep integration with Emacs org-mode for structured AI interactions
- Supports multiple AI backends (OpenAI, Azure, Anthropic, local models)
- Includes speech input/output for hands-free use
- Open source and free (GPL-3.0)
- Active community and frequent updates
- Works with both commercial and self-hosted LLMs
Cons
- Requires an OpenAI API key or other backend setup for full functionality
- Limited to Emacs users; no standalone or web interface
- Speech features require additional system configuration (e.g., Whisper, espeak)
- Image generation features depend on external services (DALL-E, Stable Diffusion)
Best For
Get AI-powered text generation and editing within Emacs org-modeCreate images from text descriptions using DALL-E or Stable DiffusionInteractively converse with ChatGPT in a structured org-mode documentAutomate project-wide AI prompts using org-ai-on-projectAdd speech-driven AI interactions in EmacsGenerate content for literate programming with noweb references
FAQ
What API keys do I need?
For OpenAI models you need an OpenAI API token. For Azure, use the Azure OpenAI endpoint. For other backends like Anthropic or local LLMs, configure accordingly.
Can I use Azure instead of OpenAI?
Yes, org-ai supports using the Azure OpenAI API instead of the default OpenAI API by configuring the org-ai-openai-api-url variable.
How do I install org-ai?
You can install org-ai via Melpa (M-x package-install org-ai), straight.el, or manually by cloning the repository and adding it to your load-path.
Does org-ai support speech input?
Yes, org-ai includes speech input/output via org-ai-talk.el. It uses Whisper for speech recognition on macOS and can be configured on other platforms.
Can I use local LLMs?
Yes, org-ai supports local LLMs through oobabooga/text-generation-webui and other local LLaMA-based models.