From 4f970c537adfb67cdd2f071d122f9a503746f38f Mon Sep 17 00:00:00 2001
From: Tyler <68524461+TySP-Dev@users.noreply.github.com>
Date: Fri, 15 May 2026 14:36:29 -0400
Subject: [PATCH] Updated README

---
 README.md | 87 +++++++++++++++++--------------------------------------
 1 file changed, 27 insertions(+), 60 deletions(-)

diff --git a/README.md b/README.md
index 4c6e075..af6768b 100644
--- a/README.md
+++ b/README.md
@@ -1,8 +1,9 @@
-# AI Answers Plugin for SearXNG  
+# Ollama AI Answers Plugin for SearXNG
 **Single file install**  
-**Does not block result loading time**  
+**Does not block result loading time**
+**Based on [ai-answers-searxng](https://github.com/cra88y/ai-answers-searxng) by [cra88y](https://github.com/cra88y)
 
-A SearXNG plugin that generates AI answers using search results as RAG context. Supports 8+ LLM providers.
+A SearXNG plugin that generates local AI overviews powered by Ollama, using search results as RAG context.
 
 Features:
 - token-by-token UI streaming
@@ -12,7 +13,6 @@ Features:
 - internally called low-latency RAG for follow ups (bypasses http loopback)
 - native network integration via `searx.network` (respects proxy/SSL settings)
 - stateless conversation persistence/sharability via URL
-- provider detection based on URL
 
 
 ## Installation
@@ -27,81 +27,48 @@ plugins:
 
 ## Configuration
 
-Configure via the environment variables:
+Configure via environment variables:
 
 ### Required
 
-- `LLM_PROVIDER`: openrouter, openai, ollama, localai, lmstudio, gemini, azure, or huggingface
-- `LLM_KEY`: Provider API key (optional for local providers: ollama, localai, lmstudio)
+- `LLM_URL`: Ollama chat completions endpoint. Default: `http://ollama:11434/v1/chat/completions`
+- `LLM_MODEL`: Model name as listed in Ollama. Default: `llama3.2`
 
 ### Optional
 
-- `LLM_MODEL`: Model identifier. Defaults vary. Recommended: 10-30B dense or 5-15B MoE activated.
-- `LLM_URL`: Overrides endpoint URL for any provider preset.
-- `LLM_SYSTEM_PROMPT`: Overrides some of the system prompt. Default `You are a direct, citation-accurate search synthesis engine.`.
-- `LLM_MAX_TOKENS`: Default `500`.
+- `LLM_SYSTEM_PROMPT`: Overrides the system prompt. Default: `You are a direct, citation-accurate search synthesis engine.`
+- `LLM_MAX_TOKENS`: Default `200`.
 - `LLM_TEMPERATURE`: Default `0.2`.
-- `LLM_CONTEXT_DEEP_COUNT`: results as context with full snippets. Default `5`.
+- `LLM_CONTEXT_DEEP_COUNT`: Results used as context with full snippets. Default `5`.
 - `LLM_CONTEXT_SHALLOW_COUNT`: Results with headlines only (additional breadth). Default `15`.
-- `LLM_TABS`: Tab whitelist, comma delimiter. Default `general,science,it,news`.
-- `LLM_INTERACTIVE`: UI mode. Default is `true` (interactive: copy, regenerate, follow up). Set to `false` for simple response only mode.
+- `LLM_TABS`: Tab whitelist, comma delimited. Default `general,science,it,news`.
+- `LLM_INTERACTIVE`: UI mode. Default `true` (interactive: copy, regenerate, follow up). Set to `false` for simple response only.
 - `LLM_QUESTION_MARK_REQUIRED`: Only trigger AI answers when the query contains `?`. Default `false`.
-- `LLM_OLLAMA_UNLOAD_AFTER`: Unload Ollama model after each response. Default `false`.
 
 ## How It Works
-1 user initial search 
-2 results return server side 
-3 `post_search` plugin hook entry
-4 token optimized context extracted 
-5 inject the ui/logic "shell" into standard results answer object 
-6 client side script calls custom endpoint with signed token
-7 LLM response streams back token by token
+1. User performs initial search
+2. Results return server side
+3. `post_search` plugin hook fires
+4. Token-optimized context extracted from results
+5. UI/logic shell injected into the standard results answer object
+6. Client-side script calls custom endpoint with a signed token
+7. Ollama response renders token by token in the UI
 
-## Examples
+## Example
 
-### OpenRouter
-```
-LLM_PROVIDER=openrouter
-LLM_KEY=sk-or-xxx
-LLM_MODEL=google/gemma-3-27b-it:free
+### Docker Compose
+```yaml
+environment:
+  - LLM_URL=http://ollama:11434/v1/chat/completions
+  - LLM_MODEL=llama3.2
 ```
 
-### Ollama (Local)
+### Environment variables
 ```
-LLM_PROVIDER=ollama
-LLM_KEY=ollama
+LLM_URL=http://ollama:11434/v1/chat/completions
 LLM_MODEL=llama3.2
 ```
 
-### LocalAI
-```
-LLM_PROVIDER=localai
-LLM_KEY=your-key
-LLM_MODEL=gpt-4
-LLM_URL=http://localai.lan:8080/v1/chat/completions
-```
-
-### Gemini
-```
-LLM_PROVIDER=gemini
-LLM_KEY=AIzaSy...
-LLM_MODEL=gemma-3-27b-it
-```
-
-### Azure
-```
-LLM_PROVIDER=azure
-LLM_KEY=your-api-key
-LLM_URL=https://your-resource.openai.azure.com/openai/deployments/your-deployment/chat/completions?api-version=2024-02-01
-```
-
-### Hugging Face
-```
-LLM_PROVIDER=huggingface
-LLM_KEY=hf_xxx
-LLM_MODEL=meta-llama/Meta-Llama-3-8B-Instruct
-```
-
 ## Development
 
 ```bash