Question 1

How do I set up org-ai to use local LLMs like CodeLlama?

Accepted Answer

Install oobabooga/text-generation-webui in a conda environment, clone the model from Hugging Face, and start the API server with --api flag, as detailed in the 'Using local LLMs' section. Then, use :local in AI blocks to send requests.

Question 2

org-ai vs gptel: which is better for ChatGPT in Emacs?

Accepted Answer

org-ai offers broader features like image generation, speech support, and project-wide operations, while gptel is more lightweight and focused on chat. Choose org-ai for multimodal needs or gptel for simpler text-based interactions.

Question 3

Can org-ai generate images offline?

Accepted Answer

Yes, by setting up Stable Diffusion locally via stable-diffusion-webui and configuring org-ai-sd-endpoint-base, you can generate images without an internet connection, though it requires significant local resources and setup.

Question 4

How to enable speech input with org-ai on Windows?

Accepted Answer

The README notes that whisper.el doesn't work out-of-the-box on Windows; you may need to patch the code or use alternatives, as mentioned in the Windows-specific steps, which involve manual workarounds.

Question 5

Does org-ai support Claude AI from Anthropic?

Accepted Answer

Yes, org-ai supports Anthropic Claude by setting the service to 'anthropic' and providing an API key, with model options like claude-3-opus, as described in the 'Using other services' section.

Question 6

Is org-ai free to use with OpenAI API?

Accepted Answer

No, using OpenAI API incurs costs based on usage; org-ai provides a command to check usage, and you need an API token, but local models can avoid fees, though they require setup and resources.

org-ai

What is org-ai?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Frequently Asked Questions