diff --git a/CHANGELOG.md b/CHANGELOG.md
index b19abd294e6..7899233e92f 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -34,6 +34,7 @@ Docs: https://docs.openclaw.ai
 - Telegram: add `channels.telegram.mediaGroupFlushMs` at the top level and per account so operators can tune album buffering instead of being stuck with the hard-coded 500ms media-group flush window. Fixes #76149. Thanks @vincentkoc.
 - Config/messages: coerce boolean `messages.visibleReplies` and `messages.groupChat.visibleReplies` values to the documented enum modes so an intuitive toggle no longer invalidates config and drops channel startup. Fixes #75390. Thanks @scottgl9.
 - Agents/network: allow trusted web-search providers and configured model-provider hosts to work behind Surge/Clash/sing-box fake-IP DNS by accepting RFC 2544 and IPv6 ULA synthetic answers only for the request's scoped hostname, without broad private-network access. Refs #76530 and #76549. Thanks @zqchris.
+- Providers: honor env-proxy settings for guarded provider model fetches when no explicit dispatcher policy is configured, preserving explicit transport overrides. Fixes #70453. (#72480) Thanks @mjamiv.
 - Feishu: accept and honor `channels.feishu.blockStreaming` at the top level and per account, while keeping the legacy default off so Feishu cards no longer reject documented config or silently drop block replies. Fixes #75555. Thanks @vincentkoc.
 - Google Chat: normalize custom Google auth transport headers before google-auth/gaxios interceptors run, restoring webhook token verification when certificate retrieval expects Fetch `Headers`. Fixes #76742. Thanks @donbowman.
 - Doctor/plugins: reset stale `plugins.slots.memory` and `plugins.slots.contextEngine` references during `doctor --fix`, so cleanup of missing plugin config does not leave unrecoverable slot owners behind. Fixes #76550 and #76551. Thanks @vincentkoc.
diff --git a/src/agents/provider-transport-fetch.test.ts b/src/agents/provider-transport-fetch.test.ts
index 5043888b43e..5228fc323ce 100644
--- a/src/agents/provider-transport-fetch.test.ts
+++ b/src/agents/provider-transport-fetch.test.ts
@@ -3,24 +3,40 @@ import { Stream } from "openai/streaming";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 
 const {
+  buildProviderRequestDispatcherPolicyMock,
   fetchWithSsrFGuardMock,
   mergeModelProviderRequestOverridesMock,
   resolveProviderRequestPolicyConfigMock,
+  shouldUseEnvHttpProxyForUrlMock,
+  withTrustedEnvProxyGuardedFetchModeMock,
 } = vi.hoisted(() => ({
+  buildProviderRequestDispatcherPolicyMock: vi.fn<
+    (_request?: unknown) => { mode: "direct" } | undefined
+  >(() => undefined),
   fetchWithSsrFGuardMock: vi.fn(),
   mergeModelProviderRequestOverridesMock: vi.fn((current, overrides) => ({
     ...current,
     ...overrides,
   })),
   resolveProviderRequestPolicyConfigMock: vi.fn(() => ({ allowPrivateNetwork: false })),
+  shouldUseEnvHttpProxyForUrlMock: vi.fn(() => false),
+  withTrustedEnvProxyGuardedFetchModeMock: vi.fn((params: Record<string, unknown>) => ({
+    ...params,
+    mode: "trusted_env_proxy",
+  })),
 }));
 
 vi.mock("../infra/net/fetch-guard.js", () => ({
   fetchWithSsrFGuard: fetchWithSsrFGuardMock,
+  withTrustedEnvProxyGuardedFetchMode: withTrustedEnvProxyGuardedFetchModeMock,
+}));
+
+vi.mock("../infra/net/proxy-env.js", () => ({
+  shouldUseEnvHttpProxyForUrl: shouldUseEnvHttpProxyForUrlMock,
 }));
 
 vi.mock("./provider-request-config.js", () => ({
-  buildProviderRequestDispatcherPolicy: vi.fn(() => ({ mode: "direct" })),
+  buildProviderRequestDispatcherPolicy: buildProviderRequestDispatcherPolicyMock,
   getModelProviderRequestTransport: vi.fn(() => undefined),
   mergeModelProviderRequestOverrides: mergeModelProviderRequestOverridesMock,
   resolveProviderRequestPolicyConfig: resolveProviderRequestPolicyConfigMock,
@@ -33,10 +49,13 @@ describe("buildGuardedModelFetch", () => {
       finalUrl: "https://api.openai.com/v1/responses",
       release: vi.fn(async () => undefined),
     });
+    buildProviderRequestDispatcherPolicyMock.mockClear().mockReturnValue(undefined);
     mergeModelProviderRequestOverridesMock.mockClear();
     resolveProviderRequestPolicyConfigMock
       .mockClear()
       .mockReturnValue({ allowPrivateNetwork: false });
+    shouldUseEnvHttpProxyForUrlMock.mockClear().mockReturnValue(false);
+    withTrustedEnvProxyGuardedFetchModeMock.mockClear();
     delete process.env.OPENCLAW_DEBUG_PROXY_ENABLED;
     delete process.env.OPENCLAW_DEBUG_PROXY_URL;
     delete process.env.OPENCLAW_SDK_RETRY_MAX_WAIT_SECONDS;
@@ -137,6 +156,63 @@ describe("buildGuardedModelFetch", () => {
     });
   });
 
+  it("uses trusted env-proxy mode for provider calls when no explicit dispatcher policy is configured", async () => {
+    shouldUseEnvHttpProxyForUrlMock.mockReturnValueOnce(true);
+    const { buildGuardedModelFetch } = await import("./provider-transport-fetch.js");
+    const model = {
+      id: "gpt-5.4",
+      provider: "openai",
+      api: "openai-responses",
+      baseUrl: "https://api.openai.com/v1",
+    } as unknown as Model<"openai-responses">;
+
+    const fetcher = buildGuardedModelFetch(model);
+    await fetcher("https://api.openai.com/v1/responses", { method: "POST" });
+
+    expect(shouldUseEnvHttpProxyForUrlMock).toHaveBeenCalledWith(
+      "https://api.openai.com/v1/responses",
+    );
+    expect(withTrustedEnvProxyGuardedFetchModeMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        url: "https://api.openai.com/v1/responses",
+        dispatcherPolicy: undefined,
+        policy: {
+          allowRfc2544BenchmarkRange: true,
+          allowIpv6UniqueLocalRange: true,
+          hostnameAllowlist: ["api.openai.com"],
+        },
+      }),
+    );
+    expect(fetchWithSsrFGuardMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        url: "https://api.openai.com/v1/responses",
+        mode: "trusted_env_proxy",
+      }),
+    );
+  });
+
+  it("keeps explicit provider dispatcher policies in strict guarded-fetch mode", async () => {
+    shouldUseEnvHttpProxyForUrlMock.mockReturnValueOnce(true);
+    buildProviderRequestDispatcherPolicyMock.mockReturnValueOnce({ mode: "direct" });
+    const { buildGuardedModelFetch } = await import("./provider-transport-fetch.js");
+    const model = {
+      id: "gpt-5.4",
+      provider: "openai",
+      api: "openai-responses",
+      baseUrl: "https://api.openai.com/v1",
+    } as unknown as Model<"openai-responses">;
+
+    const fetcher = buildGuardedModelFetch(model);
+    await fetcher("https://api.openai.com/v1/responses", { method: "POST" });
+
+    expect(withTrustedEnvProxyGuardedFetchModeMock).not.toHaveBeenCalled();
+    expect(fetchWithSsrFGuardMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        dispatcherPolicy: { mode: "direct" },
+      }),
+    );
+  });
+
   it("threads explicit transport timeouts into the shared guarded fetch seam", async () => {
     const { buildGuardedModelFetch } = await import("./provider-transport-fetch.js");
     const model = {
diff --git a/src/agents/provider-transport-fetch.ts b/src/agents/provider-transport-fetch.ts
index df4a4627c3b..055802e0af1 100644
--- a/src/agents/provider-transport-fetch.ts
+++ b/src/agents/provider-transport-fetch.ts
@@ -1,5 +1,9 @@
 import type { Api, Model } from "@mariozechner/pi-ai";
-import { fetchWithSsrFGuard } from "../infra/net/fetch-guard.js";
+import {
+  fetchWithSsrFGuard,
+  withTrustedEnvProxyGuardedFetchMode,
+} from "../infra/net/fetch-guard.js";
+import { shouldUseEnvHttpProxyForUrl } from "../infra/net/proxy-env.js";
 import {
   ssrfPolicyFromHttpBaseUrlFakeIpHostnameAllowlist,
   type SsrFPolicy,
@@ -340,7 +344,7 @@ export function buildGuardedModelFetch(model: Model<Api>, timeoutMs?: number): t
         signal: request.signal,
         ...(request.body ? ({ duplex: "half" } as const) : {}),
       } satisfies RequestInit & { duplex?: "half" });
-    const result = await fetchWithSsrFGuard({
+    const guardedFetchOptions = {
       url,
       init: requestInit ?? init,
       capture: {
@@ -356,7 +360,12 @@ export function buildGuardedModelFetch(model: Model<Api>, timeoutMs?: number): t
       // replays unsafe request bodies across cross-origin redirects.
       allowCrossOriginUnsafeRedirectReplay: false,
       ...(policy ? { policy } : {}),
-    });
+    };
+    const result = await fetchWithSsrFGuard(
+      !dispatcherPolicy && shouldUseEnvHttpProxyForUrl(url)
+        ? withTrustedEnvProxyGuardedFetchMode(guardedFetchOptions)
+        : guardedFetchOptions,
+    );
     let response = result.response;
     if (shouldBypassLongSdkRetry(response)) {
       const headers = new Headers(response.headers);