diff --git a/docs/METRICS.md b/docs/METRICS.md
index 5162479b..47202dd0 100644
--- a/docs/METRICS.md
+++ b/docs/METRICS.md
@@ -12,11 +12,15 @@ The repository includes a ready-to-run monitoring stack:
 - `prometheus.yaml` scrapes the vLLM server from the Docker host every five
   seconds.
 - `prometheus-rules.yaml` records one-minute averages for prompt, generation,
-  and total token throughput.
+  and total token throughput, plus speculative-decoding metrics (acceptance
+  rate, mean accepted length, draft and accepted token rates, and
+  per-position acceptance).
 - `grafana/provisioning/` configures Prometheus as Grafana's default data source
   and loads dashboards from disk.
 - `grafana/dashboards/vllm-throughput.json` defines the default **vLLM
   Throughput** dashboard.
+- `grafana/dashboards/vllm-spec-decode.json` defines the **vLLM Speculative
+  Decoding** dashboard.
 
 The default configuration assumes vLLM is listening on port `8000` on the
 Docker host. Confirm the endpoint first:
@@ -88,16 +92,55 @@ imported into Grafana:
 
 ## Speculative decoding and MTP panels
 
-The general vLLM dashboard may not include speculative-decoding panels. Add
-Grafana panels with the following PromQL queries.
+The provisioned **vLLM Speculative Decoding** dashboard
+(`grafana/dashboards/vllm-spec-decode.json`) covers these metrics out of the
+box. It appears in the **vLLM** folder and queries the `vllm:spec_decode_*`
+recording rules from the `vllm-spec-decode` group in `prometheus-rules.yaml`:
+
+- **Draft acceptance rate** — fraction of draft tokens accepted
+  (`vllm:spec_decode_acceptance_rate:rate1m`)
+- **Mean accepted length** — tokens emitted per verification step, including
+  the bonus token (`vllm:spec_decode_mean_accepted_length:rate1m`); with N
+  speculative tokens it ranges from 1.0 to N+1
+- **Draft tokens / second** (`vllm:spec_decode_draft_tokens_per_second:rate1m`)
+- **Draft vs accepted tokens / second** — the gap between the two series is
+  wasted draft compute
+- **Acceptance rate by draft position** — per-position quality of the MTP
+  cascade (`vllm:spec_decode_acceptance_rate_by_pos:rate1m`)
+
+The stat panels query the recorded metrics as instant values. The ratio
+rules (acceptance rate, mean accepted length, per-position) evaluate to no
+value once traffic stops, so those panels show **No value** shortly after
+the last request, when the last recorded sample leaves Prometheus'
+five-minute lookback window; this avoids a misleading 100% acceptance rate
+at idle. The token-rate rules have no such guard and keep recording zero
+while vLLM is running, so the draft and accepted token panels show 0 at
+idle.
+
+If you add or replace dashboard files on disk, restart the Grafana container
+for the file provider to reload them:
+
+```bash
+docker compose -f docker-compose.metrics.yaml restart grafana
+```
+
+If you prefer to build your own panels directly on the raw counters (for
+example with a longer rate window), the following PromQL queries are
+equivalent to the recording rules:
 
 ### Draft-token acceptance rate
 
 ```promql
-100 *
-sum(rate(vllm:spec_decode_num_accepted_tokens_total[5m]))
-/
-sum(rate(vllm:spec_decode_num_draft_tokens_total[5m]))
+(
+  100 *
+  sum(rate(vllm:spec_decode_num_accepted_tokens_total[5m]))
+  /
+  sum(rate(vllm:spec_decode_num_draft_tokens_total[5m]))
+)
+and on()
+(
+  sum(rate(vllm:spec_decode_num_draft_tokens_total[5m])) > 0
+)
 ```
 
 Use a Gauge or Time series visualization with the unit set to percent.
@@ -108,10 +151,16 @@ This convention includes the target/bonus token emitted by a verification
 step:
 
 ```promql
-1 +
-sum(rate(vllm:spec_decode_num_accepted_tokens_total[5m]))
-/
-sum(rate(vllm:spec_decode_num_drafts_total[5m]))
+(
+  1 +
+  sum(rate(vllm:spec_decode_num_accepted_tokens_total[5m]))
+  /
+  sum(rate(vllm:spec_decode_num_drafts_total[5m]))
+)
+and on()
+(
+  sum(rate(vllm:spec_decode_num_drafts_total[5m])) > 0
+)
 ```
 
 ### Draft and accepted tokens per second
@@ -131,18 +180,26 @@ sum(rate(vllm:spec_decode_num_accepted_tokens_total[5m]))
 ### Acceptance rate by draft position
 
 ```promql
-100 *
-sum by (position) (
-  rate(vllm:spec_decode_num_accepted_tokens_per_pos_total[5m])
+(
+  100 *
+  sum by (position) (
+    rate(vllm:spec_decode_num_accepted_tokens_per_pos_total[5m])
+  )
+  /
+  scalar(
+    sum(rate(vllm:spec_decode_num_drafts_total[5m]))
+  )
 )
-/
-scalar(
+and on()
+(
   sum(rate(vllm:spec_decode_num_drafts_total[5m]))
+  > 0
 )
 ```
 
-Use a Bar chart or Time series visualization and set the legend to
-`Position {{position}}`.
+The provisioned dashboard renders this metric as a Bar gauge with one bar
+per draft position. If you build your own panel, use a Bar gauge or Time
+series visualization and set the legend to `Position {{position}}`.
 
 ## Generate test traffic
 
diff --git a/grafana/dashboards/vllm-spec-decode.json b/grafana/dashboards/vllm-spec-decode.json
new file mode 100644
index 00000000..5ac99081
--- /dev/null
+++ b/grafana/dashboards/vllm-spec-decode.json
@@ -0,0 +1,284 @@
+{
+  "annotations": {
+    "list": []
+  },
+  "editable": true,
+  "fiscalYearStartMonth": 0,
+  "graphTooltip": 1,
+  "id": null,
+  "links": [],
+  "liveNow": false,
+  "panels": [
+    {
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "fieldConfig": {
+        "defaults": {
+          "decimals": 1,
+          "max": 1,
+          "min": 0,
+          "unit": "percentunit"
+        },
+        "overrides": []
+      },
+      "gridPos": {
+        "h": 6,
+        "w": 8,
+        "x": 0,
+        "y": 0
+      },
+      "id": 1,
+      "options": {
+        "colorMode": "value",
+        "graphMode": "none",
+        "justifyMode": "auto",
+        "orientation": "auto",
+        "reduceOptions": {
+          "calcs": [
+            "lastNotNull"
+          ],
+          "fields": "",
+          "values": false
+        },
+        "showPercentChange": false,
+        "textMode": "auto",
+        "wideLayout": true
+      },
+      "pluginVersion": "11.0.0",
+      "targets": [
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_acceptance_rate:rate1m",
+          "legendFormat": "Accepted",
+          "range": false,
+          "refId": "A"
+        }
+      ],
+      "title": "Draft acceptance rate (1-minute average)",
+      "type": "stat"
+    },
+    {
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "fieldConfig": {
+        "defaults": {
+          "decimals": 2,
+          "min": 1,
+          "unit": "short"
+        },
+        "overrides": []
+      },
+      "gridPos": {
+        "h": 6,
+        "w": 8,
+        "x": 8,
+        "y": 0
+      },
+      "id": 2,
+      "options": {
+        "colorMode": "value",
+        "graphMode": "none",
+        "justifyMode": "auto",
+        "orientation": "auto",
+        "reduceOptions": {
+          "calcs": [
+            "lastNotNull"
+          ],
+          "fields": "",
+          "values": false
+        },
+        "showPercentChange": false,
+        "textMode": "auto",
+        "wideLayout": true
+      },
+      "pluginVersion": "11.0.0",
+      "targets": [
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_mean_accepted_length:rate1m",
+          "legendFormat": "Tokens / step",
+          "range": false,
+          "refId": "A"
+        }
+      ],
+      "title": "Mean accepted length (incl. bonus token)",
+      "type": "stat"
+    },
+    {
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "fieldConfig": {
+        "defaults": {
+          "decimals": 1,
+          "min": 0,
+          "unit": "suffix: tokens/s"
+        },
+        "overrides": []
+      },
+      "gridPos": {
+        "h": 6,
+        "w": 8,
+        "x": 16,
+        "y": 0
+      },
+      "id": 3,
+      "options": {
+        "colorMode": "value",
+        "graphMode": "none",
+        "justifyMode": "auto",
+        "orientation": "auto",
+        "reduceOptions": {
+          "calcs": [
+            "lastNotNull"
+          ],
+          "fields": "",
+          "values": false
+        },
+        "showPercentChange": false,
+        "textMode": "auto",
+        "wideLayout": true
+      },
+      "pluginVersion": "11.0.0",
+      "targets": [
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_draft_tokens_per_second:rate1m",
+          "legendFormat": "Draft",
+          "range": false,
+          "refId": "A"
+        }
+      ],
+      "title": "Draft tokens / second",
+      "type": "stat"
+    },
+    {
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "fieldConfig": {
+        "defaults": {
+          "decimals": 1,
+          "min": 0,
+          "unit": "suffix: tokens/s"
+        },
+        "overrides": []
+      },
+      "gridPos": {
+        "h": 10,
+        "w": 24,
+        "x": 0,
+        "y": 6
+      },
+      "id": 4,
+      "options": {
+        "legend": {
+          "calcs": [
+            "mean",
+            "lastNotNull"
+          ],
+          "displayMode": "table",
+          "placement": "bottom",
+          "showLegend": true
+        },
+        "tooltip": {
+          "mode": "multi",
+          "sort": "desc"
+        }
+      },
+      "pluginVersion": "11.0.0",
+      "targets": [
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_draft_tokens_per_second:rate1m",
+          "legendFormat": "Draft tokens",
+          "range": true,
+          "refId": "A"
+        },
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_accepted_tokens_per_second:rate1m",
+          "legendFormat": "Accepted tokens",
+          "range": true,
+          "refId": "B"
+        }
+      ],
+      "title": "Draft vs accepted tokens / second (1-minute average)",
+      "type": "timeseries"
+    },
+    {
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "fieldConfig": {
+        "defaults": {
+          "decimals": 1,
+          "max": 1,
+          "min": 0,
+          "unit": "percentunit"
+        },
+        "overrides": []
+      },
+      "gridPos": {
+        "h": 10,
+        "w": 24,
+        "x": 0,
+        "y": 16
+      },
+      "id": 5,
+      "options": {
+        "displayMode": "gradient",
+        "legend": {
+          "calcs": [],
+          "displayMode": "list",
+          "placement": "bottom",
+          "showLegend": false
+        },
+        "namePlacement": "auto",
+        "orientation": "auto",
+        "showUnfilled": true,
+        "sizing": "auto",
+        "valueMode": "color"
+      },
+      "pluginVersion": "11.0.0",
+      "targets": [
+        {
+          "editorMode": "code",
+          "expr": "vllm:spec_decode_acceptance_rate_by_pos:rate1m",
+          "legendFormat": "Position {{position}}",
+          "range": false,
+          "refId": "A"
+        }
+      ],
+      "title": "Acceptance rate by draft position (1-minute average)",
+      "type": "bargauge"
+    }
+  ],
+  "refresh": "5s",
+  "schemaVersion": 39,
+  "tags": [
+    "vllm",
+    "prometheus",
+    "spec-decode"
+  ],
+  "templating": {
+    "list": []
+  },
+  "time": {
+    "from": "now-15m",
+    "to": "now"
+  },
+  "timepicker": {},
+  "timezone": "browser",
+  "title": "vLLM Speculative Decoding",
+  "uid": "vllm-spec-decode",
+  "version": 3,
+  "weekStart": ""
+}
diff --git a/prometheus-rules.yaml b/prometheus-rules.yaml
index 788996a1..0864d65e 100644
--- a/prometheus-rules.yaml
+++ b/prometheus-rules.yaml
@@ -13,3 +13,51 @@ groups:
           sum(rate(vllm:generation_tokens_total{job="vllm"}[1m]))
           +
           sum(rate(vllm:prompt_tokens_total{job="vllm"}[1m]))
+
+  - name: vllm-spec-decode
+    interval: 5s
+    rules:
+      - record: vllm:spec_decode_acceptance_rate:rate1m
+        expr: >-
+          (
+            sum(rate(vllm:spec_decode_num_accepted_tokens_total{job="vllm"}[1m]))
+            /
+            sum(rate(vllm:spec_decode_num_draft_tokens_total{job="vllm"}[1m]))
+          )
+          and on()
+          (
+            sum(rate(vllm:spec_decode_num_draft_tokens_total{job="vllm"}[1m])) > 0
+          )
+
+      - record: vllm:spec_decode_mean_accepted_length:rate1m
+        expr: >-
+          (
+            1 +
+            sum(rate(vllm:spec_decode_num_accepted_tokens_total{job="vllm"}[1m]))
+            /
+            sum(rate(vllm:spec_decode_num_drafts_total{job="vllm"}[1m]))
+          )
+          and on()
+          (
+            sum(rate(vllm:spec_decode_num_drafts_total{job="vllm"}[1m])) > 0
+          )
+
+      - record: vllm:spec_decode_draft_tokens_per_second:rate1m
+        expr: sum(rate(vllm:spec_decode_num_draft_tokens_total{job="vllm"}[1m]))
+
+      - record: vllm:spec_decode_accepted_tokens_per_second:rate1m
+        expr: sum(rate(vllm:spec_decode_num_accepted_tokens_total{job="vllm"}[1m]))
+
+      - record: vllm:spec_decode_acceptance_rate_by_pos:rate1m
+        expr: >-
+          (
+            sum by (position) (
+              rate(vllm:spec_decode_num_accepted_tokens_per_pos_total{job="vllm"}[1m])
+            )
+            /
+            scalar(sum(rate(vllm:spec_decode_num_drafts_total{job="vllm"}[1m])))
+          )
+          and on()
+          (
+            sum(rate(vllm:spec_decode_num_drafts_total{job="vllm"}[1m])) > 0
+          )
diff --git a/recipes/qwen3.8-27b-nvfp4.yaml b/recipes/qwen3.8-27b-nvfp4.yaml
index 2bf409df..c6e7a155 100644
--- a/recipes/qwen3.8-27b-nvfp4.yaml
+++ b/recipes/qwen3.8-27b-nvfp4.yaml
@@ -1,6 +1,8 @@
 # Recipe: Qwen3.8-27B-NVFP4
 # Unsloth's mixed-precision NVFP4 quant of Qwen/Qwen3.8-27B.
 # A single DGX Spark has enough unified memory for the native 262K context.
+# Native MTP requires vLLM with the Qwen GDN speculative-decoding fixes from
+# vllm-project/vllm#51812 and #51674 (merged 2026-08-14 or later).
 
 recipe_version: "1"
 name: Qwen3.8-27B-NVFP4
@@ -24,6 +26,7 @@ defaults:
   max_model_len: 262144
   max_num_seqs: 4
   max_num_batched_tokens: 8192
+  num_speculative_tokens: 3
 
 env: {}
 
@@ -41,6 +44,7 @@ command: |
     --load-format fastsafetensors \
     --enable-chunked-prefill \
     --enable-prefix-caching \
+    --speculative-config '{{"method":"mtp","num_speculative_tokens":{num_speculative_tokens}}}' \
     --enable-auto-tool-choice \
     --tool-call-parser qwen3_xml \
     --reasoning-parser qwen3