diff --git a/cmd/spinloop/serve_daemon_test.go b/cmd/spinloop/serve_daemon_test.go
index 4203f86..c597ccd 100644
--- a/cmd/spinloop/serve_daemon_test.go
+++ b/cmd/spinloop/serve_daemon_test.go
@@ -1002,7 +1002,7 @@ port = 8000
 [qwen]
 model               = Youssofal/Qwen3.8-27B-MTPLX-Optimized-Speed
 context-window      = 32768
-scheduler-mode      = parallel
+scheduler-mode      = ar_batch
 max-active-requests = 4
 model-id            = preset-alias
 `
@@ -1051,7 +1051,7 @@ func TestNodeDeployConfigMtplxFromPreset(t *testing.T) {
 	}
 	// The operator's own settings survive: the bind, and the scheduling mode,
 	// which spinloop does not compute.
-	for _, want := range []string{"--host 0.0.0.0", "--port 8000", "--scheduler-mode parallel"} {
+	for _, want := range []string{"--host 0.0.0.0", "--port 8000", "--scheduler-mode ar_batch"} {
 		if !strings.Contains(args, want) {
 			t.Errorf("a node's serve args should keep %q, got: %s", want, args)
 		}
diff --git a/cmd/spinloop/serve_test.go b/cmd/spinloop/serve_test.go
index 348c9ed..b4a6eeb 100644
--- a/cmd/spinloop/serve_test.go
+++ b/cmd/spinloop/serve_test.go
@@ -97,7 +97,7 @@ port = 8000
 [qwen]
 model               = Youssofal/Qwen3.8-27B-MTPLX-Optimized-Speed
 context-window      = 32768
-scheduler-mode      = parallel
+scheduler-mode      = ar_batch
 max-active-requests = 4
 c                   = should-stay-literal
 `
@@ -810,7 +810,7 @@ func TestCmdServe_MTPPLXPresetKeysPassThrough(t *testing.T) {
 	for _, want := range []string{
 		"--model Youssofal/Qwen3.8-27B-MTPLX-Optimized-Speed",
 		"--context-window 32768",
-		"--scheduler-mode parallel",
+		"--scheduler-mode ar_batch",
 		"--max-active-requests 4",
 		"-c should-stay-literal",
 		"--download",
diff --git a/docs/commands/serve.md b/docs/commands/serve.md
index 7b75c66..a54d826 100644
--- a/docs/commands/serve.md
+++ b/docs/commands/serve.md
@@ -230,10 +230,12 @@ BASEURL  http://127.0.0.1:8000/v1                       # mtplx serve --host/--p
 - `BASEURL` sets the bind address. With none, no bind flag is emitted and
   MTPLX's own defaults stand.
 
-The scheduling mode (`--scheduler-mode`: `serial`, `parallel`, `concurrent`) is
-per-deployment tuning, not a Spinloop field — set it in a `PRESET`, written in
-MTPLX's own long-form flags. `serve` passes every other preset key through
-unchanged, so a preset is portable only to MTPLX, as with every engine.
+The scheduling mode (`--scheduler-mode`) is per-deployment tuning, not a
+Spinloop field — set it in a `PRESET`, written in MTPLX's own long-form flags.
+`serve` passes the value through without checking it, so it must be valid for
+the installed `mtplx` (`mtplx serve --help` lists the current modes). Every
+preset key is passed through unchanged, so a preset is portable only to MTPLX,
+as with every engine.
 
 `serve` never passes `--api-key`, for the same reason as [oMLX](#omlx): it
 prints the command it runs, and a key on the line would be in your screen and
diff --git a/examples/mtplx/qwen3.8-27b/README.md b/examples/mtplx/qwen3.8-27b/README.md
index f629f08..340f1be 100644
--- a/examples/mtplx/qwen3.8-27b/README.md
+++ b/examples/mtplx/qwen3.8-27b/README.md
@@ -30,7 +30,7 @@ mtplx serve \
   --model-id qwen3.8-27b \
   --context-window 32768 \
   --max-active-requests 4 \
-  --scheduler-mode parallel \
+  --scheduler-mode ar_batch \
   --download \
   --host 127.0.0.1 --port 8000
 ```
@@ -44,8 +44,10 @@ What the flags do:
 - `--context-window` — the context a single request gets. Never scaled by
   `PARALLEL`.
 - `--max-active-requests` — an admission cap on how many requests run at once.
-- `--scheduler-mode` — `serial`, `parallel`, or `concurrent`. This is
-  per-deployment tuning, not a Spinloop field, so it lives in the
+- `--scheduler-mode` — how admitted requests execute. `spinloop` passes the
+  value through without checking it, so it must be valid for your `mtplx`
+  build (`mtplx serve --help` lists the current modes). This is per-deployment
+  tuning, not a Spinloop field, so it lives in the
   [`preset.ini`](preset.ini).
 - `--host`/`--port` — the OpenAI-compatible API is served at
   `http://127.0.0.1:8000/v1`.
diff --git a/examples/mtplx/qwen3.8-27b/preset.ini b/examples/mtplx/qwen3.8-27b/preset.ini
index b5b5ab9..56b7d61 100644
--- a/examples/mtplx/qwen3.8-27b/preset.ini
+++ b/examples/mtplx/qwen3.8-27b/preset.ini
@@ -15,4 +15,4 @@ port = 8000
 model               = Youssofal/Qwen3.8-27B-MTPLX-Optimized-Speed
 context-window      = 32768   # keep in step with the Spinloop's CONTEXT
 max-active-requests = 4       # the Spinloop's PARALLEL, when it states one
-scheduler-mode      = parallel # serial | parallel | concurrent — per-deployment
+scheduler-mode      = ar_batch # per-deployment; 'mtplx serve --help' lists the current modes