# Discourse Chatbot 🤖 **URL:** \ **Category:** Plugin\ **Tags:** ai\ **Created:** [March 1, 2023, 11:08am UTC](https://meta.discourse.org/t/discourse-chatbot/256652 "2023-03-01T11:08:25Z")\ **Posts on this page:** 1\ **Showing post:** 1128
**Author:** ![merefield](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/merefield/32/176214_2.png) [@merefield](https://meta.discourse.org/u/merefield)\ **Post date:** [August 29, 2026, 1:16pm UTC](https://meta.discourse.org/t/discourse-chatbot/256652/1128 "2026-08-29T13:16:13Z")
This PR expands Discourse Chatbot’s existing OpenAI integration with first-class support for Anthropic, Google Gemini, and xAI. It adds provider-aware credentials and model selection, with independent providers for LLM replies, embeddings, vision, and image generation. Settings show only the options relevant to the selected provider, while custom model names and URLs continue to support Azure, Ollama, proxies, and other compatible services. The PR also refreshes the documentation and marks this significant expansion as version 3.0.0. Consider the new provider support as slightly experimental, so feedback is welcome and I will look to resolve any obvious issues. > > > Previously, Chatbot coupled chat, embeddings, vision, and image generation to Op…enAI settings and sent OpenAI-specific fields to every custom endpoint, so Gemini rejected requests and alternate providers could not reliably use auxiliary models. > > This change adds shared credentials and independent provider and model selectors for LLMs, embeddings, vision, and images across OpenAI, Anthropic, Google Gemini, and xAI, including native 1,536-dimensional Gemini embeddings, provider-specific vision requests, Gemini and xAI image generation, xAI JSON image editing, and hidden capability-specific settings while preserving custom model, URL, and Azure overrides. @Jagster --- _[View the full topic](https://meta.discourse.org/t/discourse-chatbot/256652)._