diff --git a/systems/x86_64-linux/halo/models.ini b/systems/x86_64-linux/halo/models.ini
index 9d23606..83b160b 100644
--- a/systems/x86_64-linux/halo/models.ini
+++ b/systems/x86_64-linux/halo/models.ini
@@ -2,6 +2,7 @@ version = 1
 
 [*]
 flash-attn          = on
+parallel            = 1
 jinja               = true
 n-gpu-layers        = 99
 threads             = 8
@@ -20,13 +21,15 @@ spec-type           = draft-mtp
 
 [Qwen3.6-35B-A3B]
 hf                  = unsloth/Qwen3.6-35B-A3B-MTP-GGUF:UD-Q8_K_XL
-spec-draft-n-max    = 3
+spec-draft-n-max    = 2
 parallel            = 1
-c                   = 262144
+c                   = 131072
+load-on-startup     = true
 
 [Qwen3.6-27B]
 hf                  = unsloth/Qwen3.6-27B-MTP-GGUF:Q6_K
 spec-draft-n-max    = 6
 parallel            = 2
 c                   = 524288
+load-on-startup     = true