diff --git a/lib/ai/providers/volcano-vision.ts b/lib/ai/providers/volcano-vision.ts
index f7fc4c781..da81fe5e9 100644
--- a/lib/ai/providers/volcano-vision.ts
+++ b/lib/ai/providers/volcano-vision.ts
@@ -15,6 +15,83 @@ type VolcanoUrlRewriteConfig = {
 };
 
 let cachedRewriteConfig: VolcanoUrlRewriteConfig | null | undefined;
+const VOLCANO_IMAGE_FETCH_TIMEOUT_MS = 20_000;
+const VOLCANO_INLINE_MAX_SIDE = 1024;
+
+function guessImageMimeTypeFromUrl(url: string): string {
+  const cleanUrl = url.split("?")[0]?.toLowerCase() || "";
+  if (cleanUrl.endsWith(".png")) return "image/png";
+  if (cleanUrl.endsWith(".webp")) return "image/webp";
+  if (cleanUrl.endsWith(".gif")) return "image/gif";
+  if (cleanUrl.endsWith(".bmp")) return "image/bmp";
+  return "image/jpeg";
+}
+
+function shouldRetryVolcanoWithInlineImage(errorText: string): boolean {
+  if (!errorText) return false;
+  return /timeout while downloading url|download multimodal file timed out|invalidparameter/i.test(
+    errorText
+  );
+}
+
+async function fetchImageAsDataUri(url: string): Promise<string | null> {
+  if (!/^https?:\/\//i.test(url)) return null;
+
+  const controller = new AbortController();
+  const timeoutId = setTimeout(() => controller.abort(), VOLCANO_IMAGE_FETCH_TIMEOUT_MS);
+
+  try {
+    const response = await fetch(url, { signal: controller.signal });
+    if (!response.ok) {
+      console.warn("[Volcano Vision] Failed to fetch source image for inline retry:", response.status, url);
+      return null;
+    }
+
+    const arrayBuffer = await response.arrayBuffer();
+    const buffer = Buffer.from(arrayBuffer);
+    if (buffer.length === 0) return null;
+
+    const inputMime = response.headers.get("content-type") || guessImageMimeTypeFromUrl(url);
+
+    try {
+      const sharp = await import("sharp");
+      const optimized = await sharp.default(buffer)
+        .rotate()
+        .resize(VOLCANO_INLINE_MAX_SIDE, VOLCANO_INLINE_MAX_SIDE, {
+          fit: "inside",
+          withoutEnlargement: true,
+        })
+        .jpeg({ quality: 86, mozjpeg: true })
+        .toBuffer();
+      return `data:image/jpeg;base64,${optimized.toString("base64")}`;
+    } catch {
+      return `data:${inputMime};base64,${buffer.toString("base64")}`;
+    }
+  } catch (error) {
+    console.warn("[Volcano Vision] Inline image fetch failed:", error);
+    return null;
+  } finally {
+    clearTimeout(timeoutId);
+  }
+}
+
+async function convertImageFieldToInlineDataUri(
+  image: string | string[] | undefined
+): Promise<string | string[] | undefined> {
+  if (!image) return image;
+  const list = Array.isArray(image) ? image : [image];
+  const converted = await Promise.all(
+    list.map(async (item) => {
+      if (typeof item !== "string") return null;
+      if (item.startsWith("data:image")) return item;
+      return fetchImageAsDataUri(item);
+    })
+  );
+
+  const ready = converted.filter((item): item is string => typeof item === "string" && item.length > 0);
+  if (ready.length === 0) return undefined;
+  return Array.isArray(image) ? ready : ready[0];
+}
 
 async function getVolcanoUrlRewriteConfig(): Promise<VolcanoUrlRewriteConfig | null> {
   if (cachedRewriteConfig !== undefined) {
@@ -426,49 +503,66 @@ async function sendVolcanoRequest(
     }
   }
 
-  // 设置 10 分钟超时
-  const controller = new AbortController();
-  const timeoutId = setTimeout(() => controller.abort(), 600000);
+  const executeRequest = async (payload: any, tag: string): Promise<VolcanoVisionResponse> => {
+    const controller = new AbortController();
+    const timeoutId = setTimeout(() => controller.abort(), 600000);
+
+    try {
+      const response = await fetch(url, {
+        method: "POST",
+        headers: {
+          "Content-Type": "application/json",
+          Authorization: `Bearer ${apiKey}`,
+        },
+        body: JSON.stringify(payload),
+        signal: controller.signal,
+      });
+
+      if (!response.ok) {
+        const errorText = await response.text();
+        console.error(`[Volcano Vision] Request failed (${tag}):`, response.status, errorText);
+        const err = new Error(
+          `Volcano Vision API error: ${response.status} ${response.statusText} - ${errorText}`
+        ) as Error & { status?: number; body?: string };
+        err.status = response.status;
+        err.body = errorText;
+        throw err;
+      }
+
+      const data = await response.json();
+      console.log(
+        `[Volcano Vision] Response received (${tag}):`,
+        JSON.stringify(data).substring(0, 200) + "..."
+      );
+
+      if (data.error) {
+        throw new Error(`API Error: ${data.error.message || data.error.code}`);
+      }
+
+      return data as VolcanoVisionResponse;
+    } finally {
+      clearTimeout(timeoutId);
+    }
+  };
 
   try {
-    const response = await fetch(url, {
-      method: "POST",
-      headers: {
-        "Content-Type": "application/json",
-        Authorization: `Bearer ${apiKey}`,
-      },
-      body: JSON.stringify(requestBody),
-      signal: controller.signal,
-    });
-
-    clearTimeout(timeoutId);
-
-    if (!response.ok) {
-      const errorText = await response.text();
-      console.error("[Volcano Vision] Request failed:", response.status, errorText);
-      throw new Error(
-        `Volcano Vision API error: ${response.status} ${response.statusText}`
-      );
-    }
-
-    const data = await response.json();
-    console.log(
-      "[Volcano Vision] Response received:",
-      JSON.stringify(data).substring(0, 200) + "..."
-    );
-
-    // 检查是否有错误
-    if (data.error) {
-      throw new Error(`API Error: ${data.error.message || data.error.code}`);
-    }
-
-    return data as VolcanoVisionResponse;
+    return await executeRequest(requestBody, "primary");
   } catch (error) {
-    clearTimeout(timeoutId);
     if (error instanceof Error && error.name === "AbortError") {
       console.error("[Volcano Vision] Request timeout after 10 minutes");
       throw new Error("请求超时，图片处理时间过长，请稍后重试");
     }
+
+    const errorBody = (error as any)?.body || String(error || "");
+    if (requestBody.image && shouldRetryVolcanoWithInlineImage(errorBody)) {
+      console.warn("[Volcano Vision] Retrying with inline base64 image payload...");
+      const inlineImage = await convertImageFieldToInlineDataUri(requestBody.image);
+      if (inlineImage) {
+        const retryBody = { ...requestBody, image: inlineImage };
+        return await executeRequest(retryBody, "inline-retry");
+      }
+      console.warn("[Volcano Vision] Inline retry skipped: failed to convert source image(s).");
+    }
     throw error;
   }
 }