diff --git a/deployments/ai-stack/docker-compose.yml b/deployments/ai-stack/docker-compose.yml
index 311f3dc..109d9d2 100644
--- a/deployments/ai-stack/docker-compose.yml
+++ b/deployments/ai-stack/docker-compose.yml
@@ -60,8 +60,12 @@ services:
       - ollama-data:/root/.ollama
     environment:
       - OLLAMA_HOST=0.0.0.0:11434
-      - OLLAMA_KEEP_ALIVE=30m
-      - OLLAMA_MAX_LOADED_MODELS=2
+      # KEEP_ALIVE=-1 holds loaded models in VRAM until evicted by another
+      # load (vs the default 5m / our previous 30m which forces a reload
+      # penalty on every cold use). Pair with MAX_LOADED_MODELS sized to
+      # whatever fits in your GPU's VRAM — see README "VRAM sizing".
+      - OLLAMA_KEEP_ALIVE=-1
+      - OLLAMA_MAX_LOADED_MODELS=3
       - OLLAMA_FLASH_ATTENTION=1
     deploy:
       resources: