diff --git a/.vscode/settings.json b/.vscode/settings.json
index ba0a285fa8..6a914c5401 100644
--- a/.vscode/settings.json
+++ b/.vscode/settings.json
@@ -34,7 +34,7 @@
     // make stylelint work with tsx antd-style css template string
     "typescriptreact"
   ],
-  "vitest.maximumConfigs": 10,
+  "vitest.maximumConfigs": 20,
   "workbench.editor.customLabels.patterns": {
     "**/app/**/[[]*[]]/[[]*[]]/page.tsx": "${dirname(2)}/${dirname(1)}/${dirname} • page component",
     "**/app/**/[[]*[]]/page.tsx": "${dirname(1)}/${dirname} • page component",
@@ -81,8 +81,7 @@
     "**/src/store/*/slices/*/reducer.ts": "${dirname(2)}/${dirname} • reducer",
 
     "**/src/config/modelProviders/*.ts": "${filename} • provider",
-    "**/src/config/aiModels/*.ts": "${filename} • model",
-    "**/src/config/paramsSchemas/*/*.json": "${dirname(1)}/${filename} • params",
+    "**/packages/model-bank/src/aiModels/aiModels/*.ts": "${filename} • model",
     "**/packages/model-runtime/src/*/index.ts": "${dirname} • runtime",
 
     "**/src/server/services/*/index.ts": "${dirname} • server/service",
diff --git a/packages/database/src/models/__tests__/generationBatch.test.ts b/packages/database/src/models/__tests__/generationBatch.test.ts
index 45b0b39312..18dd4f8297 100644
--- a/packages/database/src/models/__tests__/generationBatch.test.ts
+++ b/packages/database/src/models/__tests__/generationBatch.test.ts
@@ -2,7 +2,7 @@
 import { eq } from 'drizzle-orm';
 import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest';
 
-import { LobeChatDatabase } from '../../type';import { AsyncTaskStatus } from '@/types/asyncTask';
+import { AsyncTaskStatus } from '@/types/asyncTask';
 import { GenerationConfig } from '@/types/generation';
 
 import {
@@ -12,6 +12,7 @@ import {
   generations,
   users,
 } from '../../schemas';
+import { LobeChatDatabase } from '../../type';
 import { GenerationBatchModel } from '../generationBatch';
 import { getTestDB } from './_util';
 
@@ -367,6 +368,51 @@ describe('GenerationBatchModel', () => {
       });
     });
 
+    it('should transform single config imageUrl through FileService', async () => {
+      const [createdBatch] = await serverDB
+        .insert(generationBatches)
+        .values({
+          ...testBatch,
+          userId,
+          config: { imageUrl: 'single-image.jpg', prompt: 'test prompt' },
+        })
+        .returning();
+
+      const results = await generationBatchModel.queryGenerationBatchesByTopicIdWithGenerations(
+        testTopic.id,
+      );
+
+      expect(results[0].config).toEqual({
+        imageUrl: 'https://example.com/single-image.jpg',
+        prompt: 'test prompt',
+      });
+    });
+
+    it('should transform both imageUrl and imageUrls when both are present', async () => {
+      const [createdBatch] = await serverDB
+        .insert(generationBatches)
+        .values({
+          ...testBatch,
+          userId,
+          config: {
+            imageUrl: 'single-image.jpg',
+            imageUrls: ['url1.jpg', 'url2.jpg'],
+            prompt: 'test prompt',
+          },
+        })
+        .returning();
+
+      const results = await generationBatchModel.queryGenerationBatchesByTopicIdWithGenerations(
+        testTopic.id,
+      );
+
+      expect(results[0].config).toEqual({
+        imageUrl: 'https://example.com/single-image.jpg',
+        imageUrls: ['https://example.com/url1.jpg', 'https://example.com/url2.jpg'],
+        prompt: 'test prompt',
+      });
+    });
+
     it('should handle config without imageUrls', async () => {
       const [createdBatch] = await serverDB
         .insert(generationBatches)
diff --git a/packages/database/src/models/generationBatch.ts b/packages/database/src/models/generationBatch.ts
index 68afad6831..c7e82f0672 100644
--- a/packages/database/src/models/generationBatch.ts
+++ b/packages/database/src/models/generationBatch.ts
@@ -1,7 +1,6 @@
 import debug from 'debug';
 import { and, eq } from 'drizzle-orm';
 
-import { LobeChatDatabase } from '../type';
 import { FileService } from '@/server/services/file';
 import { Generation, GenerationAsset, GenerationBatch, GenerationConfig } from '@/types/generation';
 
@@ -11,6 +10,7 @@ import {
   NewGenerationBatch,
   generationBatches,
 } from '../schemas/generation';
+import { LobeChatDatabase } from '../type';
 import { GenerationModel } from './generation';
 
 const log = debug('lobe-image:generation-batch-model');
@@ -121,6 +121,13 @@ export class GenerationBatchModel {
           // Transform config
           (async () => {
             const config = batch.config as GenerationConfig;
+
+            // Handle single imageUrl
+            if (config.imageUrl) {
+              config.imageUrl = await this.fileService.getFullFileUrl(config.imageUrl);
+            }
+
+            // Handle imageUrls array
             if (Array.isArray(config.imageUrls)) {
               config.imageUrls = await Promise.all(
                 config.imageUrls.map((url) => this.fileService.getFullFileUrl(url)),
diff --git a/packages/model-bank/src/aiModels/aihubmix.ts b/packages/model-bank/src/aiModels/aihubmix.ts
index 1e9771348e..1c993a4a11 100644
--- a/packages/model-bank/src/aiModels/aihubmix.ts
+++ b/packages/model-bank/src/aiModels/aihubmix.ts
@@ -700,7 +700,7 @@ const aihubmixModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description: 'Gemini 2.5 Flash 实验模型，支持图像生成',
-    displayName: 'Gemini 2.5 Flash Image Preview',
+    displayName: 'Nano Banana',
     id: 'gemini-2.5-flash-image-preview',
     maxOutput: 8192,
     pricing: {
diff --git a/packages/model-bank/src/aiModels/google.ts b/packages/model-bank/src/aiModels/google.ts
index 5c9a591277..06b9377ab0 100644
--- a/packages/model-bank/src/aiModels/google.ts
+++ b/packages/model-bank/src/aiModels/google.ts
@@ -196,8 +196,8 @@ const googleChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description:
-      'Gemini 2.5 Flash Image Preview 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
-    displayName: 'Gemini 2.5 Flash Image Preview',
+      'Nano Banana 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
+    displayName: 'Nano Banana',
     enabled: true,
     id: 'gemini-2.5-flash-image-preview',
     maxOutput: 8192,
@@ -610,12 +610,12 @@ const imagenBaseParameters: ModelParamsSchema = {
 /* eslint-disable sort-keys-fix/sort-keys-fix */
 const googleImageModels: AIImageModelCard[] = [
   {
-    displayName: 'Gemini 2.5 Flash Image Preview',
+    displayName: 'Nano Banana',
     id: 'gemini-2.5-flash-image-preview:image',
     enabled: true,
     type: 'image',
     description:
-      'Gemini 2.5 Flash Image Preview 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
+      'Nano Banana 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
     releasedAt: '2025-08-26',
     parameters: CHAT_MODEL_IMAGE_GENERATION_PARAMS,
     pricing: {
diff --git a/packages/model-bank/src/aiModels/openrouter.ts b/packages/model-bank/src/aiModels/openrouter.ts
index e25f85e3de..6ffd708b1b 100644
--- a/packages/model-bank/src/aiModels/openrouter.ts
+++ b/packages/model-bank/src/aiModels/openrouter.ts
@@ -37,7 +37,7 @@ const openrouterChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description: 'Gemini 2.5 Flash 实验模型，支持图像生成',
-    displayName: 'Gemini 2.5 Flash Image Preview',
+    displayName: 'Nano Banana',
     id: 'google/gemini-2.5-flash-image-preview',
     maxOutput: 8192,
     pricing: {
@@ -57,7 +57,7 @@ const openrouterChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description: 'Gemini 2.5 Flash 实验模型，支持图像生成',
-    displayName: 'Gemini 2.5 Flash Image Preview (free)',
+    displayName: 'Nano Banana (free)',
     id: 'google/gemini-2.5-flash-image-preview:free',
     maxOutput: 8192,
     releasedAt: '2025-08-26',
diff --git a/packages/model-bank/src/aiModels/vertexai.ts b/packages/model-bank/src/aiModels/vertexai.ts
index a8f6841f32..5d7ce95e6c 100644
--- a/packages/model-bank/src/aiModels/vertexai.ts
+++ b/packages/model-bank/src/aiModels/vertexai.ts
@@ -126,8 +126,8 @@ const vertexaiChatModels: AIChatModelCard[] = [
     },
     contextWindowTokens: 32_768 + 8192,
     description:
-      'Gemini 2.5 Flash Image Preview 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
-    displayName: 'Gemini 2.5 Flash Image Preview',
+      'Nano Banana 是 Google 最新、最快、最高效的原生多模态模型，它允许您通过对话生成和编辑图像。',
+    displayName: 'Nano Banana',
     enabled: true,
     id: 'gemini-2.5-flash-image-preview',
     maxOutput: 8192,
diff --git a/packages/model-runtime/src/google/createImage.ts b/packages/model-runtime/src/google/createImage.ts
index 3aa2ebc203..4651f45775 100644
--- a/packages/model-runtime/src/google/createImage.ts
+++ b/packages/model-runtime/src/google/createImage.ts
@@ -6,6 +6,40 @@ import { parseGoogleErrorMessage } from '../utils/googleErrorParser';
 import { imageUrlToBase64 } from '../utils/imageToBase64';
 import { parseDataUri } from '../utils/uriParser';
 
+// Maximum number of images allowed for processing
+const MAX_IMAGE_COUNT = 10;
+
+/**
+ * Process a single image URL and convert it to Google AI Part format
+ */
+async function processImageForParts(imageUrl: string): Promise<Part> {
+  const { mimeType, base64, type } = parseDataUri(imageUrl);
+
+  if (type === 'base64') {
+    if (!base64) {
+      throw new TypeError("Image URL doesn't contain base64 data");
+    }
+
+    return {
+      inlineData: {
+        data: base64,
+        mimeType: mimeType || 'image/png',
+      },
+    };
+  } else if (type === 'url') {
+    const { base64: urlBase64, mimeType: urlMimeType } = await imageUrlToBase64(imageUrl);
+
+    return {
+      inlineData: {
+        data: urlBase64,
+        mimeType: urlMimeType,
+      },
+    };
+  } else {
+    throw new TypeError(`currently we don't support image url: ${imageUrl}`);
+  }
+}
+
 /**
  * Extract image data from generateContent response
  */
@@ -71,36 +105,30 @@ async function generateImageByChatModel(
   const { model, params } = payload;
   const actualModel = model.replace(':image', '');
 
+  // Check for conflicting image parameters
+  if (params.imageUrl && params.imageUrls && params.imageUrls.length > 0) {
+    throw new TypeError('Cannot provide both imageUrl and imageUrls parameters simultaneously');
+  }
+
   // Build content parts
   const parts: Part[] = [{ text: params.prompt }];
 
   // Add image for editing if provided
   if (params.imageUrl && params.imageUrl !== null) {
-    const { mimeType, base64, type } = parseDataUri(params.imageUrl);
+    const imagePart = await processImageForParts(params.imageUrl);
+    parts.push(imagePart);
+  }
 
-    if (type === 'base64') {
-      if (!base64) {
-        throw new TypeError("Image URL doesn't contain base64 data");
-      }
-
-      parts.push({
-        inlineData: {
-          data: base64,
-          mimeType: mimeType || 'image/png',
-        },
-      });
-    } else if (type === 'url') {
-      const { base64: urlBase64, mimeType: urlMimeType } = await imageUrlToBase64(params.imageUrl);
-
-      parts.push({
-        inlineData: {
-          data: urlBase64,
-          mimeType: urlMimeType,
-        },
-      });
-    } else {
-      throw new TypeError(`currently we don't support image url: ${params.imageUrl}`);
+  // Add multiple images for editing if provided
+  if (params.imageUrls && Array.isArray(params.imageUrls) && params.imageUrls.length > 0) {
+    if (params.imageUrls.length > MAX_IMAGE_COUNT) {
+      throw new TypeError(`Too many images provided. Maximum ${MAX_IMAGE_COUNT} images allowed`);
     }
+
+    const imageParts = await Promise.all(
+      params.imageUrls.map((imageUrl) => processImageForParts(imageUrl)),
+    );
+    parts.push(...imageParts);
   }
 
   const contents: Content[] = [