diff --git a/docs/CONFIG.md b/docs/CONFIG.md
index 6205c68..b7e6836 100644
--- a/docs/CONFIG.md
+++ b/docs/CONFIG.md
@@ -78,6 +78,8 @@ net_jobs = "auto"      # auto or integer
 ```
 
 ## LLM (Ollama)
+Install Ollama: [ollama.com](https://ollama.com)
+
 ```toml
 [llm]
 mode = "off"           # off|parse|assist
@@ -89,6 +91,11 @@ max_tokens = 256
 Recommended models:
 - `Qwen2.5:latest` (safe default)
 - `Qwen2.5:14b` (higher accuracy, more RAM/VRAM)
+Alternatives:
+- `llama3.1:8b` (strong general-purpose)
+- `gemma2:9b` (fast, efficient)
+- `mistral-nemo:12b` (mid-size, long context)
+- `phi3:14b` (compact, good structure)
 
 ## Provider base URLs (for testing)
 ```toml