From 6bad214c3d9ee50d965fe86c41f1d90d0cc08a5a Mon Sep 17 00:00:00 2001
From: Your Name <your.email@example.com>
Date: Wed, 18 Feb 2026 19:37:04 +0800
Subject: [PATCH] feat: improve video mode toolbar and agent video analysis

---
 components/canvas/CanvasChatPanel.tsx |  77 +++--
 lib/ai/smart-analyzer.ts              | 452 ++++++++++++++++++++++++--
 2 files changed, 486 insertions(+), 43 deletions(-)

diff --git a/components/canvas/CanvasChatPanel.tsx b/components/canvas/CanvasChatPanel.tsx
index a49d87218..dcd61eed5 100644
--- a/components/canvas/CanvasChatPanel.tsx
+++ b/components/canvas/CanvasChatPanel.tsx
@@ -1160,8 +1160,28 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
   
   // 所有可选模型（Agent模式 + 可视化模型）
   const allModels = [SMART_AUTO_MODEL, ...visualModels];
+  const videoOnlyModels = useMemo(
+    () => visualModels.filter((model) => VIDEO_PROVIDER_TYPES.has(model.providerType)),
+    [visualModels]
+  );
   const currentModel = allModels.find((model) => model.id === selectedModelId);
   const isVideoModel = Boolean(currentModel && VIDEO_PROVIDER_TYPES.has(currentModel.providerType));
+  const selectableModels = useMemo(() => {
+    if (isDetailGenerationMode) {
+      return [SMART_AUTO_MODEL];
+    }
+    if (isVideoModel) {
+      return videoOnlyModels;
+    }
+    return allModels;
+  }, [SMART_AUTO_MODEL, allModels, isDetailGenerationMode, isVideoModel, videoOnlyModels]);
+  const modelSelectValue = selectableModels.some((model) => model.id === selectedModelId)
+    ? selectedModelId
+    : selectableModels[0]?.id || selectedModelId;
+  const selectDisplayModel =
+    selectableModels.find((model) => model.id === modelSelectValue) ||
+    currentModel ||
+    SMART_AUTO_MODEL;
   const videoFrameSlots = useMemo(() => {
     if (!isVideoModel || !videoParams) return [] as string[];
     if (videoParams.mode === "first-last-frame") return ["首帧", "尾帧"] as string[];
@@ -1210,6 +1230,11 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
     }
   }, [aiDetailGenerationEnabled, isDetailGenerationMode]);
 
+  useEffect(() => {
+    if (!modelSelectValue || modelSelectValue === selectedModelId) return;
+    setSelectedModelId(modelSelectValue);
+  }, [modelSelectValue, selectedModelId]);
+
   useEffect(() => {
     if (isDetailGenerationMode) {
       setAttachments([]);
@@ -2845,11 +2870,12 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
   );
   const shouldShowImageLoading = hasGenerationStart && !hasGenerationEnd && !!currentModel;
   const loadingImageCount = imageParams.imageCount || 1;
+  const isDetailEmptyState = isDetailGenerationMode && visibleMessages.length === 0;
 
   const renderModelSelectContent = () => (
     <SelectContent className="w-[238px] max-h-[340px] overflow-hidden rounded-[18px] border border-[#cfd4dc] bg-[#f3f4f6] p-1.5 shadow-[0_8px_20px_rgba(15,23,42,0.13)]">
       <div className="max-h-[312px] space-y-0.5 overflow-y-auto pr-1">
-        {allModels.map((model) => (
+        {selectableModels.map((model) => (
           <SelectItem
             key={model.id}
             value={model.id}
@@ -2868,12 +2894,15 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
                 </span>
                 <span className="truncate text-[12px] font-medium leading-none">{model.displayName}</span>
               </span>
-              {selectedModelId === model.id ? (
+              {modelSelectValue === model.id ? (
                 <span className="text-sm font-semibold text-[#1f2937]">✓</span>
               ) : null}
             </span>
           </SelectItem>
         ))}
+        {selectableModels.length === 0 ? (
+          <div className="px-2 py-2 text-xs text-[#6b7280]">暂无可用模型</div>
+        ) : null}
       </div>
     </SelectContent>
   );
@@ -3000,6 +3029,9 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
       </div>
       
       {/* 消息列表 */}
+      {isDetailEmptyState ? (
+        <div className="h-16 shrink-0" />
+      ) : (
       <ScrollArea
         className={cn(
           "flex-1 p-4",
@@ -3273,6 +3305,7 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
           <div ref={messagesEndRef} />
         </div>
       </ScrollArea>
+      )}
 
       {fullscreenPreview && createPortal(
         <div
@@ -3766,7 +3799,7 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
           
           {/* Bottom Toolbar */}
           <div className="absolute left-2 bottom-2 right-2 flex items-center gap-2">
-            <div className="flex min-w-0 flex-1 items-center gap-1.5 overflow-x-visible pr-0.5">
+            <div className="flex min-w-0 flex-1 items-center gap-1.5 pr-0.5">
               <input
                 ref={fileInputRef}
                 type="file"
@@ -3886,16 +3919,20 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
                 </Button>
               )}
 
-              <div className="shrink-0">
-                {isVideoModel ? (
-                  <VideoParamsBar
-                    modelName={currentModel?.displayName}
-                    initialParams={videoParams || undefined}
-                    onParamsChange={handleVideoParamsChange}
-                    inline={true}
-                    providerType={currentModel?.providerType}
-                  />
-                ) : (
+              {isVideoModel ? (
+                <div className="min-w-0 flex-1 overflow-x-auto scrollbar-hide">
+                  <div className="w-max pr-1">
+                    <VideoParamsBar
+                      modelName={currentModel?.displayName}
+                      initialParams={videoParams || undefined}
+                      onParamsChange={handleVideoParamsChange}
+                      inline={true}
+                      providerType={currentModel?.providerType}
+                    />
+                  </div>
+                </div>
+              ) : (
+                <div className="shrink-0">
                   <ImageParamsBar
                     modelName={currentModel?.displayName}
                     initialParams={imageParams}
@@ -3903,13 +3940,13 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
                     inline={true}
                     providerType={currentModel?.providerType}
                   />
-                )}
-              </div>
+                </div>
+              )}
             </div>
 
             <div className="flex items-center gap-2 flex-shrink-0">
               <Select
-                value={selectedModelId}
+                value={modelSelectValue}
                 onValueChange={(value) => {
                   setSelectedModelId(value);
                   if (value !== "auto" || !aiDetailGenerationEnabled) {
@@ -3919,15 +3956,15 @@ export function CanvasChatPanel({ chatId, initialMessages = [], onClose, onOpenF
               >
                 <SelectTrigger className="h-8 w-8 rounded-full border border-[#cfd4dc] bg-white p-0 justify-center [&>svg]:hidden">
                   <span className="h-6 w-6 overflow-hidden rounded-full bg-white flex items-center justify-center">
-                    {currentModel?.avatarUrl ? (
+                    {selectDisplayModel?.avatarUrl ? (
                       <img
-                        src={currentModel.avatarUrl}
-                        alt={currentModel.displayName || "模型头像"}
+                        src={selectDisplayModel.avatarUrl}
+                        alt={selectDisplayModel.displayName || "模型头像"}
                         className="h-full w-full object-cover"
                       />
                     ) : (
                       <span className="flex h-full w-full items-center justify-center text-[10px] font-medium text-[#4b5563]">
-                        {(currentModel?.displayName || "A").slice(0, 1)}
+                        {(selectDisplayModel?.displayName || "A").slice(0, 1)}
                       </span>
                     )}
                   </span>
diff --git a/lib/ai/smart-analyzer.ts b/lib/ai/smart-analyzer.ts
index b824ddc23..87f413e9f 100644
--- a/lib/ai/smart-analyzer.ts
+++ b/lib/ai/smart-analyzer.ts
@@ -6,6 +6,7 @@
  */
 
 import { generateText } from "ai";
+import sharp from "sharp";
 // 使用动态导入避免循环依赖问题
 // import { getLanguageModelById } from "./dynamic-provider";
 // import { getEnabledAiModelsWithProvider } from "@/lib/db/queries";
@@ -33,6 +34,12 @@ function inferDefaultImageEditSubIntent(text: string): AnalysisResult["subIntent
   return "enhance";
 }
 
+function isVideoGenerationRequest(text: string): boolean {
+  return /视频|动画|动起来|生成视频|做视频|视频生成|图生视频|文生视频|镜头|分镜|运镜|首帧|尾帧/.test(
+    text.toLowerCase(),
+  );
+}
+
 function applyDirectExecutionPolicy(
   result: AnalysisResult,
   params: {
@@ -42,6 +49,22 @@ function applyDirectExecutionPolicy(
     imageFromHistory?: boolean;
   }
 ): AnalysisResult {
+  if (isVideoGenerationRequest(params.text)) {
+    return {
+      ...result,
+      intent: "video_gen",
+      subIntent: undefined,
+      confidence: Math.max(result.confidence, 0.8),
+      needsClarification: false,
+      clarifyingQuestions: [],
+      conversationalReply:
+        result.conversationalReply ||
+        (params.hasImage
+          ? "收到，我先按你这轮图片做视频需求分析，再自动优化提示词执行视频生成。"
+          : "收到，我先做视频需求分析并自动优化提示词后执行生成。"),
+    };
+  }
+
   const hasUsableImage =
     params.hasImage || Boolean(params.imageFromHistory) || (params.imageUrls?.length || 0) > 0;
   if (!hasUsableImage) return result;
@@ -149,6 +172,7 @@ const ANALYZER_SYSTEM_PROMPT = `你是一个友好的AI图片编辑助手。用
 - 说明你理解了什么需求，让用户知道你在做什么
 - 有图片时优先直接执行，不要提问
  - 如果提供了画布上下文，仅用于理解，不要在 conversationalReply 中复述坐标或技术细节
+- video_gen 场景要明确说明：会先分析视频需求/分镜，再自动优化提示词并执行生成
 
 ## 示例
 用户："帮我把背景换成白色"
@@ -203,8 +227,340 @@ const ANALYZER_SYSTEM_PROMPT = `你是一个友好的AI图片编辑助手。用
   "confidence": 0.9,
   "conversationalReply": "好的，我再帮你提升一下清晰度~",
   "needsClarification": false
+}
+
+用户："把这张图做成5秒短视频，镜头慢慢推进"
+{
+  "intent": "video_gen",
+  "confidence": 0.9,
+  "conversationalReply": "收到，我先按你的目标做视频需求分析和分镜规划，再自动优化提示词执行生成~",
+  "needsClarification": false
 }`;
 
+const ANALYZER_MAX_IMAGE_COUNT = 2;
+const ANALYZER_IMAGE_MAX_SIDE = 500;
+const ANALYZER_IMAGE_FETCH_TIMEOUT_MS = 10000;
+const ZHIPU_VISION_MODEL_CANDIDATES = [
+  "glm-4.6v-flash",
+  "glm-4.6v",
+  "glm-4v-plus-0111",
+  "glm-4v-plus",
+  "glm-4v",
+];
+
+function isHttpUrl(value: string) {
+  return /^https?:\/\//i.test(value);
+}
+
+async function fetchWithTimeout(url: string, timeoutMs: number): Promise<Response> {
+  const controller = new AbortController();
+  const timeoutId = setTimeout(() => controller.abort(), timeoutMs);
+  try {
+    return await fetch(url, {
+      signal: controller.signal,
+      cache: "no-store",
+      headers: {
+        "user-agent": "ai-chatbot-smart-analyzer/1.0",
+      },
+    });
+  } finally {
+    clearTimeout(timeoutId);
+  }
+}
+
+async function toAnalyzerDataUrl(imageUrl: string): Promise<string | null> {
+  try {
+    const response = await fetchWithTimeout(imageUrl, ANALYZER_IMAGE_FETCH_TIMEOUT_MS);
+    if (!response.ok) {
+      console.warn(
+        `[Smart Analyzer] Failed to fetch image for inline payload (${response.status}): ${imageUrl}`,
+      );
+      return null;
+    }
+
+    const contentType = response.headers.get("content-type") || "";
+    if (!contentType.startsWith("image/")) {
+      console.warn(`[Smart Analyzer] Skip non-image content-type: ${contentType}`);
+      return null;
+    }
+
+    const arrayBuffer = await response.arrayBuffer();
+    const inputBuffer = Buffer.from(arrayBuffer);
+
+    const optimized = await sharp(inputBuffer)
+      .rotate()
+      .resize({
+        width: ANALYZER_IMAGE_MAX_SIDE,
+        height: ANALYZER_IMAGE_MAX_SIDE,
+        fit: "inside",
+        withoutEnlargement: true,
+      })
+      .jpeg({
+        quality: 80,
+        mozjpeg: true,
+      })
+      .toBuffer();
+
+    return `data:image/jpeg;base64,${optimized.toString("base64")}`;
+  } catch (error) {
+    console.warn("[Smart Analyzer] Failed to convert image URL to inline base64:", error);
+    return null;
+  }
+}
+
+async function buildAnalyzerVisionImages(imageUrls?: string[]) {
+  const urls = (imageUrls || []).filter((url) => isHttpUrl(url)).slice(0, ANALYZER_MAX_IMAGE_COUNT);
+  if (urls.length === 0) return [] as Array<{ type: "image"; image: string }>;
+
+  const convertedResults = await Promise.all(urls.map((url) => toAnalyzerDataUrl(url)));
+
+  return urls.map((originalUrl, index) => {
+    const inlineDataUrl = convertedResults[index];
+    return {
+      type: "image" as const,
+      image: inlineDataUrl || originalUrl,
+    };
+  });
+}
+
+function parseAnalysisResultFromModelText(
+  responseText: string,
+  userText: string,
+  hasImage: boolean,
+): AnalysisResult | null {
+  const jsonMatch = responseText.match(/\{[\s\S]*\}/);
+  if (!jsonMatch) return null;
+
+  try {
+    const result = JSON.parse(jsonMatch[0]) as AnalysisResult;
+    return normalizeIntentForImageContext(result, userText, hasImage);
+  } catch (error) {
+    console.error("[Smart Analyzer] Failed to parse model response:", error);
+    return null;
+  }
+}
+
+function isLikelyZhipuEndpoint(baseUrl?: string | null): boolean {
+  if (!baseUrl) return false;
+  return /open\.bigmodel\.cn/i.test(baseUrl);
+}
+
+function normalizeZhipuChatEndpoint(baseUrl: string): string {
+  const cleaned = baseUrl.replace(/\/+$/, "");
+  if (/\/chat\/completions$/i.test(cleaned)) return cleaned;
+  if (/\/api\/paas\/v4$/i.test(cleaned)) return `${cleaned}/chat/completions`;
+  if (/\/api\/paas\/v4\//i.test(cleaned)) return cleaned;
+  if (/open\.bigmodel\.cn/i.test(cleaned)) return `${cleaned}/api/paas/v4/chat/completions`;
+  return `${cleaned}/chat/completions`;
+}
+
+function toZhipuImageUrl(image: string): string {
+  const dataUrlPrefix = /^data:image\/[a-zA-Z0-9.+-]+;base64,/i;
+  if (dataUrlPrefix.test(image)) {
+    return image.replace(dataUrlPrefix, "");
+  }
+  return image;
+}
+
+async function analyzeWithZhipuOfficialApi(params: {
+  apiKey: string;
+  baseUrl: string;
+  userMessage: string;
+  visionImages: Array<{ type: "image"; image: string }>;
+  userText: string;
+  hasImage: boolean;
+}): Promise<AnalysisResult | null> {
+  const endpoint = normalizeZhipuChatEndpoint(params.baseUrl);
+
+  const content = [
+    { type: "text", text: params.userMessage },
+    ...params.visionImages.map((imagePart) => ({
+      type: "image_url",
+      image_url: {
+        url: toZhipuImageUrl(imagePart.image),
+      },
+    })),
+  ];
+
+  for (const modelId of ZHIPU_VISION_MODEL_CANDIDATES) {
+    try {
+      const response = await fetch(endpoint, {
+        method: "POST",
+        headers: {
+          "Content-Type": "application/json",
+          Authorization: `Bearer ${params.apiKey}`,
+        },
+        body: JSON.stringify({
+          model: modelId,
+          messages: [
+            {
+              role: "system",
+              content: ANALYZER_SYSTEM_PROMPT,
+            },
+            {
+              role: "user",
+              content,
+            },
+          ],
+          temperature: 0.1,
+          max_tokens: 360,
+        }),
+      });
+
+      const payload = await response.json().catch(() => ({}));
+      if (!response.ok || payload?.error) {
+        const errorCode = String(payload?.error?.code || "");
+        const errorMessage = String(payload?.error?.message || "");
+        if (errorCode === "1211" || /模型不存在/.test(errorMessage)) {
+          continue;
+        }
+        console.warn(
+          `[Smart Analyzer] Zhipu official vision request failed (model=${modelId})`,
+          payload?.error || payload,
+        );
+        continue;
+      }
+
+      const message = payload?.choices?.[0]?.message;
+      const responseText = String(message?.content || message?.reasoning_content || "").trim();
+      if (!responseText) {
+        continue;
+      }
+
+      const parsed = parseAnalysisResultFromModelText(
+        responseText,
+        params.userText,
+        params.hasImage,
+      );
+      if (parsed) {
+        console.log(`[Smart Analyzer] Zhipu official vision model used: ${modelId}`);
+        return parsed;
+      }
+    } catch (error) {
+      console.warn(
+        `[Smart Analyzer] Zhipu official vision request exception (model=${modelId})`,
+        error,
+      );
+    }
+  }
+
+  return null;
+}
+
+async function summarizeWithZhipuOfficialApi(params: {
+  apiKey: string;
+  baseUrl: string;
+  promptText: string;
+  visionImages: Array<{ type: "image"; image: string }>;
+}): Promise<string | null> {
+  const endpoint = normalizeZhipuChatEndpoint(params.baseUrl);
+  const content = [
+    { type: "text", text: params.promptText },
+    ...params.visionImages.map((imagePart) => ({
+      type: "image_url",
+      image_url: {
+        url: toZhipuImageUrl(imagePart.image),
+      },
+    })),
+  ];
+
+  for (const modelId of ZHIPU_VISION_MODEL_CANDIDATES) {
+    try {
+      const response = await fetch(endpoint, {
+        method: "POST",
+        headers: {
+          "Content-Type": "application/json",
+          Authorization: `Bearer ${params.apiKey}`,
+        },
+        body: JSON.stringify({
+          model: modelId,
+          messages: [
+            {
+              role: "system",
+              content:
+                "你是图像观察助手。请根据图片给出执行前说明，输出严格两行中文：第1行“画面：...”，第2行“关键要素：...”。不要输出其他内容。",
+            },
+            {
+              role: "user",
+              content,
+            },
+          ],
+          temperature: 0.2,
+          max_tokens: 140,
+        }),
+      });
+
+      const payload = await response.json().catch(() => ({}));
+      if (!response.ok || payload?.error) {
+        const errorCode = String(payload?.error?.code || "");
+        const errorMessage = String(payload?.error?.message || "");
+        if (errorCode === "1211" || /模型不存在/.test(errorMessage)) {
+          continue;
+        }
+        continue;
+      }
+
+      const message = payload?.choices?.[0]?.message;
+      const responseText = String(message?.content || message?.reasoning_content || "").trim();
+      if (responseText) {
+        console.log(`[Smart Analyzer] Zhipu official summary model used: ${modelId}`);
+        return responseText;
+      }
+    } catch {
+      continue;
+    }
+  }
+
+  return null;
+}
+
+export async function summarizeImageForExecutionReport(params: {
+  text?: string;
+  imageUrls?: string[];
+  context?: string;
+}): Promise<string | null> {
+  const { text, imageUrls, context } = params;
+  const visionImages = await buildAnalyzerVisionImages(imageUrls);
+  if (visionImages.length === 0) return null;
+
+  const {
+    getEnabledAiModelsWithProvider,
+    getAgentAiModelWithProvider,
+  } = await import("@/lib/db/queries");
+  const allModels = await getEnabledAiModelsWithProvider();
+  const agentModel = await getAgentAiModelWithProvider();
+  const zhipuProviderFromAgent =
+    agentModel?.provider?.apiKey && isLikelyZhipuEndpoint(agentModel.provider.baseUrl)
+      ? agentModel.provider
+      : null;
+  const zhipuProviderFallback = allModels.find(
+    (m) =>
+      m.provider?.apiKey &&
+      (m.provider.type === "zhipu" || isLikelyZhipuEndpoint(m.provider.baseUrl)),
+  )?.provider;
+  const zhipuProvider = zhipuProviderFromAgent || zhipuProviderFallback;
+
+  const promptText = [
+    text ? `用户需求：${text}` : "",
+    context ? `画布上下文：${context.slice(0, 220)}` : "",
+    "请给出当前画面的客观描述，便于执行前说明。",
+  ]
+    .filter(Boolean)
+    .join("\n");
+
+  if (zhipuProvider?.apiKey && zhipuProvider?.baseUrl) {
+    const summary = await summarizeWithZhipuOfficialApi({
+      apiKey: zhipuProvider.apiKey,
+      baseUrl: zhipuProvider.baseUrl,
+      promptText,
+      visionImages,
+    });
+    if (summary) return summary;
+  }
+
+  return null;
+}
+
 // 历史消息类型
 interface HistoryMessage {
   role: "user" | "assistant";
@@ -239,7 +595,7 @@ export async function analyzeIntent(params: {
     return applyDirectExecutionPolicy(quickResult, { text, hasImage, imageUrls, imageFromHistory });
   }
 
-  // 尝试使用 Qwen3 VL 进行深度分析（传入历史对话）
+  // 尝试使用多模态模型进行深度分析（传入历史对话）
   try {
     const llmResult = await analyzWithLLM(text, hasImage, imageUrls, historyMessages, context);
     if (llmResult) {
@@ -291,24 +647,45 @@ async function analyzWithLLM(
   const agentModel = await getAgentAiModelWithProvider();
   const isReasoningCompatibleProvider = (providerType: string) =>
     ["zhipu", "alibaba", "openai", "custom", "local"].includes(providerType);
+  const isLikelyZhipuVisionModel = (displayName: string, modelId: string, providerType: string) =>
+    providerType === "zhipu" ||
+    /(glm[-_.]?4v|glm[-_.]?4\.?1v|glm[-_.]?4\.?5v|vision)/i.test(
+      `${displayName} ${modelId}`,
+    );
+  const zhipuVisionModel = allModels.find(
+    (m) =>
+      m.model.supportsVision &&
+      m.model.isEnabled &&
+      !m.model.isSubModel &&
+      isLikelyZhipuVisionModel(m.model.displayName, m.model.modelId, m.provider.type),
+  );
   const qwenVisionModel = allModels.find(
     (m) =>
       m.provider.type === "alibaba" &&
       m.model.supportsVision &&
+      m.model.isEnabled &&
+      !m.model.isSubModel &&
       /(qwen|通义|vl)/i.test(`${m.model.displayName} ${m.model.modelId}`)
   );
   const fallbackVisionModel =
     qwenVisionModel ||
-    allModels.find((m) => m.provider.type === "alibaba" && m.model.supportsVision);
+    allModels.find(
+      (m) =>
+        m.provider.type === "alibaba" &&
+        m.model.supportsVision &&
+        m.model.isEnabled &&
+        !m.model.isSubModel,
+    );
+  const preferredVisionModel = zhipuVisionModel || fallbackVisionModel;
   const usableAgentModel =
     agentModel && isReasoningCompatibleProvider(agentModel.provider.type)
       ? agentModel
       : undefined;
   const selectedAnalyzerModel = hasImage
-    ? fallbackVisionModel
+    ? preferredVisionModel
     : usableAgentModel ||
       allModels.find((m) => m.provider.type === "zhipu" && !m.model.isSubModel) ||
-      fallbackVisionModel;
+      preferredVisionModel;
 
   if (!selectedAnalyzerModel) {
     console.log("[Smart Analyzer] No available analyzer model found");
@@ -348,6 +725,42 @@ ${historyText}
     ? `${contextMessage}用户上传了图片，请求内容："${text}"`
     : `${contextMessage}用户请求（无图片）："${text}"`;
 
+  const visionImages = hasImage ? await buildAnalyzerVisionImages(imageUrls) : [];
+  if (visionImages.length > 0) {
+    const inlineCount = visionImages.filter((item) => item.image.startsWith("data:image/")).length;
+    console.log(
+      `[Smart Analyzer] Prepared analyzer images: total=${visionImages.length}, inline=${inlineCount}, remote=${visionImages.length - inlineCount}`,
+    );
+  }
+
+  const zhipuProviderFromAgent =
+    usableAgentModel &&
+    usableAgentModel.provider &&
+    usableAgentModel.provider.apiKey &&
+    isLikelyZhipuEndpoint(usableAgentModel.provider.baseUrl)
+      ? usableAgentModel.provider
+      : null;
+  const zhipuProviderFallback = allModels.find(
+    (m) =>
+      m.provider?.apiKey &&
+      (m.provider.type === "zhipu" || isLikelyZhipuEndpoint(m.provider.baseUrl)),
+  )?.provider;
+  const zhipuProvider = zhipuProviderFromAgent || zhipuProviderFallback;
+
+  if (hasImage && visionImages.length > 0 && zhipuProvider?.apiKey && zhipuProvider?.baseUrl) {
+    const zhipuOfficialResult = await analyzeWithZhipuOfficialApi({
+      apiKey: zhipuProvider.apiKey,
+      baseUrl: zhipuProvider.baseUrl,
+      userMessage,
+      visionImages,
+      userText: text,
+      hasImage,
+    });
+    if (zhipuOfficialResult) {
+      return zhipuOfficialResult;
+    }
+  }
+
   console.log(
     "[Smart Analyzer] Calling analyzer model:",
     selectedAnalyzerModel.model.displayName,
@@ -356,11 +769,6 @@ ${historyText}
     historyMessages?.length || 0
   );
 
-  const visionImages = (imageUrls || [])
-    .filter((url) => /^https?:\/\//i.test(url))
-    .slice(0, 4)
-    .map((url) => ({ type: "image" as const, image: url }));
-
   // 调用模型分析（有图时携带图片，确保真实视觉分析）
   const requestPayload = hasImage && visionImages.length > 0
     ? {
@@ -388,18 +796,7 @@ ${historyText}
 
   const { text: responseText } = await generateText(requestPayload as any);
 
-  // 解析JSON响应
-  const jsonMatch = responseText.match(/\{[\s\S]*\}/);
-  if (jsonMatch) {
-    try {
-      const result = JSON.parse(jsonMatch[0]) as AnalysisResult;
-      return normalizeIntentForImageContext(result, text, hasImage);
-    } catch (e) {
-      console.error("[Smart Analyzer] Failed to parse LLM response:", e);
-    }
-  }
-
-  return null;
+  return parseAnalysisResultFromModelText(responseText, text, hasImage);
 }
 
 /**
@@ -414,6 +811,15 @@ export function fallbackKeywordAnalysis(
 
   // === 图片编辑场景（需要有图片） ===
   if (hasImage) {
+    // 视频生成（图生视频/首尾帧等）
+    if (isVideoGenerationRequest(lowerText)) {
+      return {
+        intent: "video_gen",
+        confidence: 0.9,
+        conversationalReply: "收到，我先按你提供的图片做视频需求分析和分镜规划，再自动优化提示词执行生成~",
+      };
+    }
+
     // 抠图/去背景
     if (/抠图|去背景|透明背景|去背|扣图|移除背景/.test(lowerText)) {
       return {
@@ -544,11 +950,11 @@ export function fallbackKeywordAnalysis(
   }
 
   // 视频生成
-  if (/视频|动画|动起来|生成视频|做视频|视频生成/.test(lowerText)) {
+  if (isVideoGenerationRequest(lowerText)) {
     return { 
       intent: "video_gen", 
       confidence: 0.85,
-      conversationalReply: "好的，我来帮你生成视频~",
+      conversationalReply: "好的，我先做视频需求分析并自动优化提示词后执行生成~",
     };
   }