diff --git a/admin/app/services/benchmark_service.ts b/admin/app/services/benchmark_service.ts
index 68f69b7f..a2d4e23a 100644
--- a/admin/app/services/benchmark_service.ts
+++ b/admin/app/services/benchmark_service.ts
@@ -30,6 +30,7 @@ import type {
 } from '../../types/benchmark.js'
 import KVStore from '#models/kv_store'
 import { normalizeArchitecture, deriveOsName } from '../utils/platform_metadata.js'
+import { isUnresolvedGpuModel } from '../utils/gpu_model.js'
 import { getFreeBytes } from '../utils/image_disk_preflight.js'
 import { readFile } from 'node:fs/promises'
 import { randomUUID, createHmac } from 'node:crypto'
@@ -186,35 +187,8 @@ function isSelfHostedOllamaUrl(rawUrl: string): boolean {
   return false
 }
 
-/**
- * Is this GPU "model" a placeholder rather than a real name?
- *
- * systeminformation resolves PCI ids against the container's pci.ids database.
- * When a card is newer than that file it reports the raw id verbatim — an RTX
- * 5060 comes back as "Device 2d05". Vendor detection still succeeds, so these
- * strings otherwise pass as legitimate model names and reach the leaderboard.
- *
- * Matches the "Device <hex>" shape plus the empty/unknown cases. Deliberately
- * narrow: it must never reject a real product name, and no shipping GPU is
- * called "Device" followed by four hex digits.
- */
-function isUnresolvedGpuModel(model: string): boolean {
-  const s = model.trim()
-  if (s === '') return true
-  if (/^device\s+[0-9a-f]{4}$/i.test(s)) return true
-  if (/^unknown$/i.test(s)) return true
-  // WSL2 exposes the GPU through /dev/dxg rather than the real adapter, so
-  // si.graphics() reports Microsoft's generic placeholder even while CUDA work
-  // is running on a physical card. Left unhandled, an RTX 3090 reaches the
-  // public leaderboard labelled "Microsoft Basic Render Driver", which is worse
-  // than no label: it is wrong, and it fragments per-hardware grouping (#1218).
-  //
-  // These are Microsoft's own placeholder adapter names and appear nowhere
-  // outside a Windows/WSL graphics stack, so this cannot reject a real product
-  // name on a native Linux host.
-  if (/^microsoft basic (render driver|display adapter)$/i.test(s)) return true
-  return false
-}
+// Moved to app/utils/gpu_model.ts so the Settings > System display path can
+// share the same definition — see the note there (#1196).
 
 // Minimum free disk required before pulling the AI model on first run. llama3.1:8b
 // (Q4) is ~4.9 GB; this covers the model plus headroom for the transient sysbench
diff --git a/admin/app/services/system_service.ts b/admin/app/services/system_service.ts
index d9ab50d7..846bd0f7 100644
--- a/admin/app/services/system_service.ts
+++ b/admin/app/services/system_service.ts
@@ -21,6 +21,7 @@ import env from '#start/env'
 import KVStore from '#models/kv_store'
 import { KV_STORE_SCHEMA, KVStoreKey } from '../../types/kv_store.js'
 import { isNewerVersion } from '../utils/version.js'
+import { isUnresolvedGpuModel } from '../utils/gpu_model.js'
 import { invalidateAssistantNameCache } from '../../config/inertia.js'
 import { invalidateMinRelevanceCache } from '../utils/rag_relevance.js'
 import { KiwixLibraryService } from '#services/kiwix_library_service'
@@ -501,12 +502,31 @@ export class SystemService {
           }
         }
 
-        // Run the probes when controllers are empty (common inside Docker) or
-        // when lspci gave us bogus discrete-GPU BAR0 values that need replacing.
+        // The same pci.ids staleness that produces bogus VRAM also produces a
+        // placeholder model name — a card newer than the container's pci.ids is
+        // reported as its raw id, e.g. "Device 2d05" for an RTX 5060 (#1165).
+        //
+        // These are independent symptoms, not one. lspci can hand back a
+        // perfectly plausible BAR0 reading alongside an unresolved name, and in
+        // that case nothing above fires, no probe runs, and Settings > System
+        // renders the raw PCI id as the GPU model (#1196). NVIDIA usually hides
+        // this because the nvidia-smi path tends to be reached for other
+        // reasons; AMD has no equivalent, so it shows the raw id.
+        //
+        // The probes below resolve a real name from Ollama's own startup log,
+        // so trigger them on an unresolved name too rather than only on VRAM.
+        const hasUnresolvedGpuName = (graphics.controllers || []).some((c) =>
+          isUnresolvedGpuModel(c.model || '')
+        )
+
+        // Run the probes when controllers are empty (common inside Docker),
+        // when lspci gave us bogus discrete-GPU BAR0 values that need replacing,
+        // or when it named a card it couldn't resolve.
         if (
           !graphics.controllers ||
           graphics.controllers.length === 0 ||
-          hasLspciBogusDgpuVram
+          hasLspciBogusDgpuVram ||
+          hasUnresolvedGpuName
         ) {
           const runtimes = dockerInfo.Runtimes || {}
           gpuHealth.hasNvidiaRuntime = 'nvidia' in runtimes
diff --git a/admin/app/utils/gpu_model.ts b/admin/app/utils/gpu_model.ts
new file mode 100644
index 00000000..27bc93c8
--- /dev/null
+++ b/admin/app/utils/gpu_model.ts
@@ -0,0 +1,36 @@
+/**
+ * Is this GPU "model" a placeholder rather than a real name?
+ *
+ * systeminformation resolves PCI ids against the container's pci.ids database.
+ * When a card is newer than that file it reports the raw id verbatim — an RTX
+ * 5060 comes back as "Device 2d05". Vendor detection still succeeds, so these
+ * strings otherwise pass as legitimate model names.
+ *
+ * Deliberately narrow: it must never reject a real product name. No shipping
+ * GPU is called "Device" followed by four hex digits, and the Microsoft entries
+ * below are placeholder adapter names that appear nowhere outside a Windows or
+ * WSL graphics stack.
+ *
+ * Lives in its own module rather than beside its first caller because both the
+ * leaderboard submission path and the Settings > System display path need the
+ * same answer, and they diverged once already: #1165 fixed the submission and
+ * left the System page reporting the raw id (#1196). One definition, two
+ * callers, so the next placeholder shape only has to be added once.
+ */
+export function isUnresolvedGpuModel(model: string): boolean {
+  const s = model.trim()
+  if (s === '') return true
+  if (/^device\s+[0-9a-f]{4}$/i.test(s)) return true
+  if (/^unknown$/i.test(s)) return true
+  // WSL2 exposes the GPU through /dev/dxg rather than the real adapter, so
+  // si.graphics() reports Microsoft's generic placeholder even while CUDA work
+  // is running on a physical card. Left unhandled, an RTX 3090 reaches the
+  // public leaderboard labelled "Microsoft Basic Render Driver", which is worse
+  // than no label: it is wrong, and it fragments per-hardware grouping (#1218).
+  //
+  // These are Microsoft's own placeholder adapter names and appear nowhere
+  // outside a Windows/WSL graphics stack, so this cannot reject a real product
+  // name on a native Linux host.
+  if (/^microsoft basic (render driver|display adapter)$/i.test(s)) return true
+  return false
+}
diff --git a/admin/tests/unit/gpu_model.spec.ts b/admin/tests/unit/gpu_model.spec.ts
new file mode 100644
index 00000000..651dd45f
--- /dev/null
+++ b/admin/tests/unit/gpu_model.spec.ts
@@ -0,0 +1,57 @@
+import * as assert from 'node:assert/strict'
+import { test } from 'node:test'
+
+import { isUnresolvedGpuModel } from '../../app/utils/gpu_model.js'
+
+test('unresolved PCI ids are rejected', () => {
+  // The #1165 shape: a card newer than the container's pci.ids database comes
+  // back as its raw id. "Device 2d05" is a real RTX 5060 that reached the
+  // public leaderboard under that name.
+  for (const s of ['Device 2d05', 'device 2d05', 'DEVICE 1002', 'Device  2d05']) {
+    assert.equal(isUnresolvedGpuModel(s), true, s)
+  }
+})
+
+test('empty and unknown are rejected', () => {
+  for (const s of ['', '   ', 'Unknown', 'unknown', 'UNKNOWN']) {
+    assert.equal(isUnresolvedGpuModel(s), true, JSON.stringify(s))
+  }
+})
+
+test('WSL placeholder adapter names are rejected', () => {
+  // #1218: WSL reaches the GPU via /dev/dxg, so si.graphics() reports
+  // Microsoft's placeholder while CUDA runs on a physical card.
+  for (const s of [
+    'Microsoft Basic Render Driver',
+    'microsoft basic render driver',
+    'Microsoft Basic Display Adapter',
+  ]) {
+    assert.equal(isUnresolvedGpuModel(s), true, s)
+  }
+})
+
+test('real product names are never rejected', () => {
+  // The one thing this function must not do. A false positive here discards a
+  // correct GPU name and sends "unknown" to the leaderboard instead.
+  for (const s of [
+    'NVIDIA GeForce RTX 5090',
+    'NVIDIA GeForce RTX 3090 Ti',
+    'AMD Radeon RX 6800',
+    'Navi 48 [Radeon AI PRO R9700]',
+    'Intel Iris Xe Graphics',
+    'Raphael',
+    'Phoenix1',
+    'Device Manager Graphics 5000',
+    'GeForce GTX 1080',
+    'Microsoft Basic Render Driver Pro',
+  ]) {
+    assert.equal(isUnresolvedGpuModel(s), false, s)
+  }
+})
+
+test('a hex-like id that is not the placeholder shape is kept', () => {
+  // Guard the regex against over-reach: only a bare "Device <4 hex>" qualifies.
+  for (const s of ['Device 2d05x', 'Device 2d0', 'Device 2d055', 'My Device 2d05']) {
+    assert.equal(isUnresolvedGpuModel(s), false, s)
+  }
+})