diff --git a/bin/lib/nim.js b/bin/lib/nim.js
index b735dff9632..64c074dad79 100644
--- a/bin/lib/nim.js
+++ b/bin/lib/nim.js
@@ -35,12 +35,14 @@ function detectGpu() {
       const perGpuMB = lines.map((l) => parseInt(l.trim(), 10)).filter((n) => !isNaN(n));
       if (perGpuMB.length > 0) {
         const totalMemoryMB = perGpuMB.reduce((a, b) => a + b, 0);
+        // Only mark nimCapable if at least one NIM model fits in GPU VRAM
+        const canRunNim = nimImages.models.some((m) => m.minGpuMemoryMB <= totalMemoryMB);
         return {
           type: "nvidia",
           count: perGpuMB.length,
           totalMemoryMB,
           perGpuMB: perGpuMB[0],
-          nimCapable: true,
+          nimCapable: canRunNim,
         };
       }
     }
diff --git a/bin/lib/onboard.js b/bin/lib/onboard.js
index c440cb9b75a..3c7b7f972f6 100644
--- a/bin/lib/onboard.js
+++ b/bin/lib/onboard.js
@@ -1587,6 +1587,9 @@ async function preflight() {
   const gpu = nim.detectGpu();
   if (gpu && gpu.type === "nvidia") {
     console.log(`  ✓ NVIDIA GPU detected: ${gpu.count} GPU(s), ${gpu.totalMemoryMB} MB VRAM`);
+    if (!gpu.nimCapable) {
+      console.log("  ⓘ GPU VRAM too small for local NIM — will use cloud inference");
+    }
   } else if (gpu && gpu.type === "apple") {
     console.log(`  ✓ Apple GPU detected: ${gpu.name}${gpu.cores ? ` (${gpu.cores} cores)` : ""}, ${gpu.totalMemoryMB} MB unified memory`);
     console.log("  ⓘ NIM requires NVIDIA GPU — will use cloud inference");