fix: allow private OpenAI image endpoints

CHANGELOG.md

@@ -20,6 +20,7 @@ Docs: https://docs.openclaw.ai
 - Agents/OpenAI: surface selected-model capacity failures from PI, Codex, and auto-reply harness paths with a model-switch hint instead of the generic empty-response error. Thanks @vincentkoc.
 - Providers/OpenAI: route `openai/gpt-image-2` through configured Codex OAuth directly when an `openai-codex` profile is active, instead of probing `OPENAI_API_KEY` first.
 - Providers/OpenAI: harden image generation auth routing and Codex OAuth response parsing so fallback only applies to public OpenAI API routes and bounded SSE results. Thanks @Takhoffman.
+- Providers/OpenAI: honor the private-network SSRF opt-in for OpenAI-compatible image generation endpoints, so trusted LocalAI/LAN `image_generate` routes work without disabling SSRF checks globally. Fixes #62879. Thanks @seitzbg.
 - Providers/OpenAI: stop advertising the removed `gpt-5.3-codex-spark` Codex model through fallback catalogs, and suppress stale rows with a GPT-5.5 recovery hint.
 - Plugins/QR: replace legacy `qrcode-terminal` QR rendering with bounded `qrcode-tui` helpers for plugin login/setup flows. (#65969) Thanks @vincentkoc.
 - Voice-call/realtime: wait for OpenAI session configuration before greeting or forwarding buffered audio, and reject non-allowlisted Twilio callers before stream setup. (#43501) Thanks @forrestblount.

docs/providers/openai.md

@@ -236,6 +236,10 @@ does not first try `OPENAI_API_KEY` or silently fall back to an API key for that
 request. Configure `models.providers.openai` explicitly with an API key,
 custom base URL, or Azure endpoint when you want the direct OpenAI Images API
 route instead.
+If that custom image endpoint is on a trusted LAN/private address, also set
+`browser.ssrfPolicy.dangerouslyAllowPrivateNetwork: true`; OpenClaw keeps
+private/internal OpenAI-compatible image endpoints blocked unless this opt-in is
+present.
 
 Generate:
 

docs/tools/image-generation.md

@@ -35,6 +35,10 @@ Codex OAuth uses the same `openai/gpt-image-2` model ref. When an
 through that same OAuth profile instead of first trying `OPENAI_API_KEY`.
 Explicit custom `models.providers.openai` image config, such as an API key or
 custom/Azure base URL, opts back into the direct OpenAI Images API route.
+For OpenAI-compatible LAN endpoints such as LocalAI, keep the custom
+`models.providers.openai.baseUrl` and explicitly opt in with
+`browser.ssrfPolicy.dangerouslyAllowPrivateNetwork: true`; private/internal
+image endpoints remain blocked by default.
 
 3. Ask the agent: _"Generate an image of a friendly robot mascot."_
 

extensions/openai/image-generation-provider.test.ts

@@ -316,6 +316,47 @@ describe("openai image generation provider", () => {
     expect(result.images).toHaveLength(1);
   });
 
+  it("allows OpenAI-compatible private image endpoints when browser SSRF policy opts in", async () => {
+    mockGeneratedPngResponse();
+
+    const provider = buildOpenAIImageGenerationProvider();
+    const result = await provider.generateImage({
+      provider: "openai",
+      model: "flux2-klein",
+      prompt: "A simple, clean illustration of a red apple with a green leaf",
+      cfg: {
+        browser: {
+          ssrfPolicy: {
+            dangerouslyAllowPrivateNetwork: true,
+          },
+        },
+        models: {
+          providers: {
+            openai: {
+              baseUrl: "http://192.168.1.15:8082/v1",
+              apiKey: "local-noauth",
+              models: [],
+            },
+          },
+        },
+      },
+    });
+
+    expect(resolveProviderHttpRequestConfigMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        baseUrl: "http://192.168.1.15:8082/v1",
+        allowPrivateNetwork: true,
+      }),
+    );
+    expect(postJsonRequestMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        url: "http://192.168.1.15:8082/v1/images/generations",
+        allowPrivateNetwork: true,
+      }),
+    );
+    expect(result.images).toHaveLength(1);
+  });
+
   it("forwards generation count and custom size overrides", async () => {
     mockGeneratedPngResponse();
 

extensions/openai/image-generation-provider.ts

@@ -21,6 +21,7 @@ import {
   resolveProviderHttpRequestConfig,
   sanitizeConfiguredModelProviderRequest,
 } from "openclaw/plugin-sdk/provider-http";
+import { isPrivateNetworkOptInEnabled } from "openclaw/plugin-sdk/ssrf-runtime";
 import { OPENAI_DEFAULT_IMAGE_MODEL as DEFAULT_OPENAI_IMAGE_MODEL } from "./default-models.js";
 import { resolveConfiguredOpenAIBaseUrl } from "./shared.js";
 
@@ -190,6 +191,9 @@ function shouldAllowPrivateImageEndpoint(req: {
   if (req.provider === MOCK_OPENAI_PROVIDER_ID) {
     return true;
   }
+  if (isPrivateNetworkOptInEnabled(req.cfg?.browser?.ssrfPolicy)) {
+    return true;
+  }
   const baseUrl = resolveConfiguredOpenAIBaseUrl(req.cfg);
   if (!baseUrl.startsWith("http://127.0.0.1:") && !baseUrl.startsWith("http://localhost:")) {
     return false;