feat(core): add configurable image generation models

packages/cli/src/acp-integration/acpAgent.test.ts

@@ -795,6 +795,7 @@ import type { Config } from '@qwen-code/qwen-code-core';
 import type { LoadedSettings } from '../config/settings.js';
 import type { CliArgs } from '../config/config.js';
 import {
+  AuthType,
   SessionEndReason,
   MCPServerConfig,
   SessionService,
@@ -4021,6 +4022,14 @@ describe('QwenAgent MCP SSE/HTTP support', () => {
           baseUrl: 'https://api.example.com',
           isRuntimeModel: false,
         },
+        {
+          id: 'qwen-image-2.0',
+          label: 'Qwen Image 2.0',
+          authType: 'qwen',
+          baseUrl: 'https://api.example.com',
+          imageOnly: true,
+          isRuntimeModel: false,
+        },
       ]),
       getToolRegistry: vi
         .fn()
@@ -4064,6 +4073,26 @@ describe('QwenAgent MCP SSE/HTTP support', () => {
     expect(
       preflight.cells.find((c) => c.kind === 'tool_registry')?.status,
     ).toBe('ok');
+    expect(preflight.cells.find((c) => c.kind === 'providers')).toMatchObject({
+      status: 'ok',
+      detail: { count: 1, providers: ['qwen'] },
+    });
+
+    vi.mocked(mockConfig.getAllConfiguredModels).mockReturnValue([
+      {
+        id: 'qwen-image-2.0',
+        label: 'Qwen Image 2.0',
+        authType: AuthType.QWEN_OAUTH,
+        imageOnly: true,
+      },
+    ]);
+    const imageOnlyPreflight = (await agent.extMethod(
+      SERVE_STATUS_EXT_METHODS.workspacePreflight,
+      {},
+    )) as { cells: Array<{ kind: string; status: string }> };
+    expect(
+      imageOnlyPreflight.cells.find((c) => c.kind === 'providers')?.status,
+    ).toBe('error');
 
     mockConnectionState.resolve();
     await agentPromise;

packages/cli/src/acp-integration/acpAgent.ts

@@ -5855,7 +5855,9 @@ class QwenAgent implements Agent {
 
   private buildProvidersPreflightCell(config: Config): ServePreflightCell {
     try {
-      const models = config.getAllConfiguredModels();
+      const models = config
+        .getAllConfiguredModels()
+        .filter((model) => !model.imageOnly);
       const authType = config.getAuthType?.();
       if (models.length === 0) {
         // `authType` set but zero models = the next `POST /session` will

packages/cli/src/config/config.test.ts

@@ -1327,6 +1327,24 @@ describe('loadCliConfig', () => {
     );
   });
 
+  it('should propagate the image model selection', async () => {
+    process.argv = ['node', 'script.js'];
+    const argv = await parseArguments();
+
+    await loadCliConfig(
+      {
+        imageModel: 'openai:qwen-image-2.0\0https://images.example.com/api/v1',
+      },
+      argv,
+    );
+
+    expect(mockConfigConstructorParams).toHaveBeenCalledWith(
+      expect.objectContaining({
+        imageModel: 'openai:qwen-image-2.0\0https://images.example.com/api/v1',
+      }),
+    );
+  });
+
   it('places session-injected (ACP/IDE) MCP servers at the top precedence tier', async () => {
     process.argv = ['node', 'script.js'];
     const argv = await parseArguments();

packages/cli/src/config/config.ts

@@ -2277,6 +2277,7 @@ export async function loadCliConfig(
     webSearch:
       bareMode || safeMode ? undefined : resolveWebSearchSettings(settings),
     visionModel: settings.visionModel || undefined,
+    imageModel: settings.imageModel || undefined,
     visionBridgeTimeoutMs: settings.visionBridgeTimeoutMs,
     modelFallbacks: resolveModelFallbacks(
       argv.fallbackModel,

packages/cli/src/config/settingsSchema.test.ts

@@ -35,6 +35,7 @@ describe('SettingsSchema', () => {
         'advanced',
         'plansDirectory',
         'voiceModel',
+        'imageModel',
       ];
 
       expectedSettings.forEach((setting) => {
@@ -179,6 +180,16 @@ describe('SettingsSchema', () => {
       expect(voiceModel.showInDialog).toBe(false);
     });
 
+    it('should define the image model setting', () => {
+      const imageModel = getSettingsSchema().imageModel;
+
+      expect(imageModel.type).toBe('string');
+      expect(imageModel.category).toBe('Model');
+      expect(imageModel.default).toBe('');
+      expect(imageModel.requiresRestart).toBe(false);
+      expect(imageModel.showInDialog).toBe(false);
+    });
+
     it('should define the built-in Explore model setting', () => {
       const exploreModel =
         getSettingsSchema().agents.properties.builtin.properties.exploreModel;

packages/cli/src/config/settingsSchema.ts

@@ -1309,6 +1309,17 @@ const SETTINGS_SCHEMA = {
     showInDialog: true,
   },
 
+  imageModel: {
+    type: 'string',
+    label: 'Image Model',
+    category: 'Model',
+    requiresRestart: false,
+    default: '',
+    description:
+      'Model used by the built-in image_gen tool. Set with /model --image. The selected model must be marked imageOnly in modelProviders.',
+    showInDialog: false,
+  },
+
   visionBridgeTimeoutMs: {
     type: 'integer',
     label: 'Vision Bridge Timeout (ms)',

packages/cli/src/i18n/locales/en.js

@@ -219,6 +219,7 @@ export default {
   'toolDisplayName.ExitWorktree': 'toolDisplayName.ExitWorktree',
   'toolDisplayName.Workflow': 'toolDisplayName.Workflow',
   'toolDisplayName.ReadMcpResource': 'toolDisplayName.ReadMcpResource',
+  'toolDisplayName.ImageGen': 'toolDisplayName.ImageGen',
   // ============================================================================
   // Help / UI Components
   // ============================================================================
@@ -1490,6 +1491,8 @@ export default {
     'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, [model-id] to switch immediately).',
   'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).':
     'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).',
+  'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --image for the image generation model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).':
+    'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --image for the image generation model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).',
   "Inline one-shot override isn't supported in this mode — run '/model {{model}}' first, then send your prompt.":
     "Inline one-shot override isn't supported in this mode — run '/model {{model}}' first, then send your prompt.",
   "Inline one-shot override can't switch providers. '{{model}}' belongs to a different provider — run '/model {{model}}' first, then send your prompt.":
@@ -1503,17 +1506,23 @@ export default {
     'Set the model for voice transcription',
   'Set the image-capable model used to transcribe images for a text-only main model':
     'Set the image-capable model used to transcribe images for a text-only main model',
+  'Set the model used to generate images':
+    'Set the model used to generate images',
   'Persist the model selection to the project settings (workspace scope)':
     'Persist the model selection to the project settings (workspace scope)',
   'Persist the model selection to the user settings (global scope)':
     'Persist the model selection to the user settings (global scope)',
   'Select Fast Model': 'Select Fast Model',
   'Select Vision Model': 'Select Vision Model',
+  'Select Image Model': 'Select Image Model',
   'Select Voice Model': 'Select Voice Model',
   'Vision Model': 'Vision Model',
+  'Image Model': 'Image Model',
   'Voice Model': 'Voice Model',
   'Selected voice model is unavailable.':
     'Selected voice model is unavailable.',
+  'Selected image model is unavailable.':
+    'Selected image model is unavailable.',
   "Voice model '{{model}}' is configured more than once. Remove duplicate model ids before selecting it for voice transcription.":
     "Voice model '{{model}}' is configured more than once. Remove duplicate model ids before selecting it for voice transcription.",
   'Voice dictation: {{status}} (mode: {{mode}}, {{modelText}}).':
@@ -1761,8 +1770,16 @@ export default {
     'Current voice model: {{voiceModel}}\nUse "/model --voice <model-id>" to set voice model.',
   'Current vision model: {{visionModel}}\nUse "/model --vision <model-id>" to set the vision bridge model.':
     'Current vision model: {{visionModel}}\nUse "/model --vision <model-id>" to set the vision bridge model.',
+  'Current image model: {{imageModel}}\nUse "/model --image <model-id>" to set the image generation model.':
+    'Current image model: {{imageModel}}\nUse "/model --image <model-id>" to set the image generation model.',
   "Voice model '{{modelName}}' is ambiguous. Configure a unique model id before using /model --voice.":
     "Voice model '{{modelName}}' is ambiguous. Configure a unique model id before using /model --voice.",
+  "Image model '{{modelName}}' matches multiple configured endpoints. Run /model --image without an argument and choose the exact endpoint.":
+    "Image model '{{modelName}}' matches multiple configured endpoints. Run /model --image without an argument and choose the exact endpoint.",
+  "Image model '{{modelName}}' must declare a valid HTTPS baseUrl and credential environment variable.":
+    "Image model '{{modelName}}' must declare a valid HTTPS baseUrl and credential environment variable.",
+  "'{{model}}' must declare a valid HTTPS baseUrl and credential environment variable.":
+    "'{{model}}' must declare a valid HTTPS baseUrl and credential environment variable.",
   none: 'none',
   unknown: 'unknown',
   // ============================================================================

packages/cli/src/i18n/locales/zh-TW.js

@@ -210,6 +210,7 @@ export default {
   'toolDisplayName.ExitWorktree': '退出 Worktree',
   'toolDisplayName.Workflow': '工作流程',
   'toolDisplayName.ReadMcpResource': '讀取 MCP 資源',
+  'toolDisplayName.ImageGen': '圖像生成',
 
   '↑ to manage attachments': '↑ 管理附件',
   '← → select, Delete to remove, ↓ to exit': '← → 選擇，Delete 刪除，↓ 退出',
@@ -1316,6 +1317,8 @@ export default {
     '切換此會話的模型（--fast 可設置建議模型，--voice 可設置語音轉寫模型，[model-id] 可立即切換）',
   'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).':
     '切換此會話的模型（--fast 建議模型，--voice 語音轉寫模型，--vision 視覺橋接模型，--project 持久化到專案設定，--global 持久化到使用者設定，[model-id] 立即切換，或用 [model-id] [prompt] 在另一個模型上執行一次性提示；內聯提示按原文發送，不展開 @file）',
+  'Switch the model for this session (--fast for suggestion model, --voice for voice transcription model, --vision for the vision bridge model, --image for the image generation model, --project to persist to project settings, --global to persist to user settings, [model-id] to switch immediately, or [model-id] [prompt] to run a one-off prompt on another model; the inline prompt is sent verbatim without @file expansion).':
+    '切換此會話的模型（--fast 建議模型，--voice 語音轉寫模型，--vision 視覺橋接模型，--image 圖像生成模型，--project 持久化到專案設定，--global 持久化到使用者設定，[model-id] 立即切換，或用 [model-id] [prompt] 在另一個模型上執行一次性提示；內聯提示按原文發送，不展開 @file）',
   "Inline one-shot override isn't supported in this mode — run '/model {{model}}' first, then send your prompt.":
     "此模式不支援內聯一次性覆寫——請先執行 '/model {{model}}'，再發送你的提示。",
   "Inline one-shot override can't switch providers. '{{model}}' belongs to a different provider — run '/model {{model}}' first, then send your prompt.":
@@ -1328,16 +1331,20 @@ export default {
   'Set the model for voice transcription': '設定語音轉寫模型',
   'Set the image-capable model used to transcribe images for a text-only main model':
     '設定用於為純文字主模型轉寫圖像的圖像能力模型',
+  'Set the model used to generate images': '設置用於生成圖像的模型',
   'Persist the model selection to the project settings (workspace scope)':
     '將模型選擇持久化到專案設定（工作區）',
   'Persist the model selection to the user settings (global scope)':
     '將模型選擇持久化到使用者設定（全域）',
   'Select Fast Model': '選擇快速模型',
   'Select Vision Model': '選擇視覺模型',
+  'Select Image Model': '選擇圖像模型',
   'Select Voice Model': '選擇語音模型',
   'Vision Model': '視覺模型',
+  'Image Model': '圖像模型',
   'Voice Model': '語音模型',
   'Selected voice model is unavailable.': '所選語音模型不可用。',
+  'Selected image model is unavailable.': '所選圖像模型不可用。',
   "Voice model '{{model}}' is configured more than once. Remove duplicate model ids before selecting it for voice transcription.":
     "語音模型 '{{model}}' 被配置了多次。請先移除重複的模型 ID，再將其選為語音轉寫模型。",
   'Voice dictation: {{status}} (mode: {{mode}}, {{modelText}}).':
@@ -1543,8 +1550,16 @@ export default {
     '當前語音模型：{{voiceModel}}\n使用 "/model --voice <model-id>" 設置語音模型。',
   'Current vision model: {{visionModel}}\nUse "/model --vision <model-id>" to set the vision bridge model.':
     '當前視覺模型：{{visionModel}}\n使用 "/model --vision <model-id>" 設置視覺橋接模型。',
+  'Current image model: {{imageModel}}\nUse "/model --image <model-id>" to set the image generation model.':
+    '當前圖像模型：{{imageModel}}\n使用 "/model --image <model-id>" 設置圖像生成模型。',
   "Voice model '{{modelName}}' is ambiguous. Configure a unique model id before using /model --voice.":
     "語音模型 '{{modelName}}' 不唯一。請先配置唯一的模型 ID，再使用 /model --voice。",
+  "Image model '{{modelName}}' matches multiple configured endpoints. Run /model --image without an argument and choose the exact endpoint.":
+    "圖像模型 '{{modelName}}' 匹配了多個已配置的端點。請執行 /model --image（不帶參數）並選擇確切的端點。",
+  "Image model '{{modelName}}' must declare a valid HTTPS baseUrl and credential environment variable.":
+    "圖像模型 '{{modelName}}' 必須宣告有效的 HTTPS baseUrl 和憑據環境變數。",
+  "'{{model}}' must declare a valid HTTPS baseUrl and credential environment variable.":
+    "'{{model}}' 必須宣告有效的 HTTPS baseUrl 和憑據環境變數。",
   none: '無',
   unknown: '未知',
   'Manage folder trust settings': '管理檔案夾信任設置',