doc refresh

Author: thushan

docs/content/integrations/frontend/openwebui.md

@@ -4,10 +4,16 @@ description: Connect OpenWebUI to multiple Ollama instances through Olla proxy.
 keywords: OpenWebUI, Olla, Ollama, LLM interface, web UI, load balancing, model management
 ---
 
-# OpenWebUI Integration
+# OpenWebUI Integration with Ollama
 
 OpenWebUI is a powerful web interface for interacting with LLMs. Olla acts as a proxy between OpenWebUI and your Ollama backends, providing load balancing, failover and unified model management across multiple Ollama instances.
 
+**Set in OpenWebUI:**
+
+```bash
+export OLLAMA_BASE_URL="http://localhost:40114/olla/ollama"
+```
+
 You can find an example integration of OpenWebUI with Olla and Ollama instances in <code>examples/ollama-openwebui</code> - see [latest in Github](https://github.com/thushan/olla/tree/main/examples/ollama-openwebui).
 
 ## Overview

docs/content/integrations/overview.md

@@ -26,7 +26,12 @@ You can use the `type` in [Endpoint Configurations](/olla/configuration/overview
 
 ## Frontend Support
 
-* [OpenWebUI](./frontend/openwebui.md) - native support for [OpenWebUI](https://github.com/open-webui/open-webui).
+### OpenWebUI
+
+Native support for [OpenWebUI](https://github.com/open-webui/open-webui) with Olla via:
+
+* [OpenWebUI with Ollama](./frontend/openwebui.md)
+* [OpenWebUI with OpenAI](./frontend/openwebui-openai.md)
 
 ## Profiles
 

docs/mkdocs.yml

@@ -157,7 +157,8 @@ nav:
             - Lemonade SDK: integrations/backend/lemonade.md
             - LiteLLM: integrations/backend/litellm.md
         - Frontends:
-            - OpenWebUI: integrations/frontend/openwebui.md
+            - OpenWebUI + Ollama: integrations/frontend/openwebui.md
+            - OpenWebUI + OpenAI: integrations/frontend/openwebui-openai.md
   - Development:
       - Overview: development/overview.md
       - Setup: development/setup.md

examples/ollama-openwebui/README.md

@@ -2,6 +2,8 @@
 
 This example demonstrates how to set up OpenWebUI with Olla as a proxy/load balancer for multiple Ollama instances.
 
+However, you can also opt to use any OpenAI Compatible API as well (so you can use vLLM, SgLang backends for instance).
+
 ## Architecture
 
 ```