diff --git a/CHANGELOG.md b/CHANGELOG.md
index cb65d93406b..6e1a83278e2 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -16,6 +16,7 @@ Docs: https://docs.openclaw.ai
 
 ### Fixes
 
+- CLI/model probes: fail local `infer model run` probes when the provider returns no text output, so unreachable local providers and empty completions no longer look like successful smoke tests. Refs #73023. Thanks @pavelyortho-cyber.
 - CLI/Ollama: run local `infer model run` through the lean provider completion path and skip global model discovery for one-shot local probes, so Ollama smoke tests no longer pay full chat-agent/tool startup cost or hang before the native `/api/chat` request. Fixes #72851. Thanks @TotalRes2020.
 - Daemon/service: only emit hard-coded version-manager paths such as `~/.volta/bin`, `~/.asdf/shims`, `~/.bun/bin`, and fnm/pnpm fallbacks into gateway and node service PATHs when the directories exist, so `openclaw doctor` no longer flags `gateway.path.non-minimal` against a PATH the daemon just wrote. Env-driven roots and stable user-bin dirs remain unconditional. Fixes #71944; carries forward #71964. Thanks @Sanjays2402.
 - Channels/commands: make generated `/dock-*` commands switch the active session reply route through `session.identityLinks` instead of falling through to normal chat. Fixes #69206; carries forward #73033. Thanks @clawbones and @michaelatamuk.
diff --git a/docs/cli/infer.md b/docs/cli/infer.md
index fe82af9be73..9203bad6b64 100644
--- a/docs/cli/infer.md
+++ b/docs/cli/infer.md
@@ -159,6 +159,7 @@ openclaw infer model run --local --model openai/gpt-4.1 --prompt "Reply with exa
 Notes:
 
 - Local `model run` is the narrowest CLI smoke for provider/model/auth health because it sends only the supplied prompt to the selected model.
+- Local `model run` exits non-zero when the provider returns no text output, so unreachable local providers and empty completions do not look like successful probes.
 - Use `model run --gateway` when you need to test Gateway routing, agent-runtime setup, or Gateway-managed provider state instead of the lean local completion path.
 - `model auth login`, `model auth logout`, and `model auth status` manage saved provider auth state.
 
diff --git a/src/cli/capability-cli.test.ts b/src/cli/capability-cli.test.ts
index fbadc4e7e8d..09510657a6d 100644
--- a/src/cli/capability-cli.test.ts
+++ b/src/cli/capability-cli.test.ts
@@ -419,6 +419,24 @@ describe("capability cli", () => {
     );
   });
 
+  it("fails local model probes when the provider returns no text output", async () => {
+    mocks.completeWithPreparedSimpleCompletionModel.mockResolvedValueOnce({
+      content: [],
+    } as never);
+
+    await expect(
+      runRegisteredCli({
+        register: registerCapabilityCli as (program: Command) => void,
+        argv: ["capability", "model", "run", "--prompt", "hello", "--json"],
+      }),
+    ).rejects.toThrow("exit 1");
+
+    expect(mocks.runtime.error).toHaveBeenCalledWith(
+      expect.stringContaining('No text output returned for provider "openai" model "gpt-5.4"'),
+    );
+    expect(mocks.runtime.writeJson).not.toHaveBeenCalled();
+  });
+
   it("runs gateway model probes without chat-agent prompt policy or tools", async () => {
     await runRegisteredCli({
       register: registerCapabilityCli as (program: Command) => void,
diff --git a/src/cli/capability-cli.ts b/src/cli/capability-cli.ts
index d72fa71c48f..a79ad3164d8 100644
--- a/src/cli/capability-cli.ts
+++ b/src/cli/capability-cli.ts
@@ -112,7 +112,7 @@ type CapabilityEnvelope = {
 const CAPABILITY_METADATA: CapabilityMetadata[] = [
   {
     id: "model.run",
-    description: "Run a one-shot text inference turn through the agent runtime.",
+    description: "Run a one-shot text inference turn through the selected model provider.",
     transports: ["local", "gateway"],
     flags: ["--prompt", "--model", "--local", "--gateway", "--json"],
     resultShape: "normalized payloads plus provider/model attribution",
@@ -570,6 +570,13 @@ function requireProviderModelOverride(
   };
 }
 
+function collectModelRunText(content: Array<{ type: string; text?: string }>): string {
+  return content
+    .map((block) => (block.type === "text" && typeof block.text === "string" ? block.text : ""))
+    .join("")
+    .trim();
+}
+
 async function runModelRun(params: {
   prompt: string;
   model?: string;
@@ -607,10 +614,12 @@ async function runModelRun(params: {
             : undefined,
       },
     });
-    const text = result.content
-      .map((block) => (block.type === "text" ? block.text : ""))
-      .join("")
-      .trim();
+    const text = collectModelRunText(result.content);
+    if (!text) {
+      throw new Error(
+        `No text output returned for provider "${prepared.selection.provider}" model "${prepared.selection.modelId}".`,
+      );
+    }
     return {
       ok: true,
       capability: "model.run",
@@ -618,14 +627,12 @@ async function runModelRun(params: {
       provider: prepared.selection.provider,
       model: prepared.selection.modelId,
       attempts: [],
-      outputs: text
-        ? [
-            {
-              text,
-              mediaUrl: null,
-            },
-          ]
-        : [],
+      outputs: [
+        {
+          text,
+          mediaUrl: null,
+        },
+      ],
     } satisfies CapabilityEnvelope;
   }