fix(cli): keep /model switches session-scoped

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

@@ -1040,6 +1040,8 @@ export default {
   // ============================================================================
   'Switch the model for this session (--fast for suggestion model, [model-id] to switch immediately).':
     'Switch the model for this session (--fast for suggestion model, [model-id] to switch immediately).',
+  'Switch the model for this session (--default to persist, --fast for suggestion model).':
+    'Switch the model for this session (--default to persist, --fast for suggestion model).',
   'Set a lighter model for prompt suggestions and speculative execution':
     'Set a lighter model for prompt suggestions and speculative execution',
   'Content generator configuration not available.':

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

@@ -884,6 +884,8 @@ export default {
     '生成摘要失敗 - 未從 LLM 響應中接收到文本內容',
   'Switch the model for this session (--fast for suggestion model, [model-id] to switch immediately).':
     '切換此會話的模型（--fast 可設置建議模型）',
+  'Switch the model for this session (--default to persist, --fast for suggestion model).':
+    '切換此會話的模型（--default 可持久化，--fast 可設置建議模型）',
   'Set a lighter model for prompt suggestions and speculative execution':
     '設置用於輸入建議和推測執行的輕量模型',
   'Content generator configuration not available.': '內容生成器配置不可用',

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

@@ -991,6 +991,8 @@ export default {
   // ============================================================================
   'Switch the model for this session (--fast for suggestion model, [model-id] to switch immediately).':
     '切换此会话的模型（--fast 可设置建议模型）',
+  'Switch the model for this session (--default to persist, --fast for suggestion model).':
+    '切换此会话的模型（--default 可持久化，--fast 可设置建议模型）',
   'Set a lighter model for prompt suggestions and speculative execution':
     '设置用于输入建议和推测执行的轻量模型',
   'Content generator configuration not available.': '内容生成器配置不可用',

packages/cli/src/ui/commands/modelCommand.test.ts

@@ -45,8 +45,9 @@ describe('modelCommand', () => {
   it('should have the correct name and description', () => {
     expect(modelCommand.name).toBe('model');
     expect(modelCommand.description).toBe(
-      'Switch the model for this session (--fast for suggestion model, [model-id] to switch immediately).',
+      'Switch the model for this session (--default to persist, --fast for suggestion model).',
     );
+    expect(modelCommand.argumentHint).toBe('[--default|--fast] [<model-id>]');
   });
 
   it('should return error when config is not available', async () => {
@@ -151,7 +152,7 @@ describe('modelCommand', () => {
     });
   });
 
-  it('should switch the main model directly in interactive mode when args are provided', async () => {
+  it('should switch the main model for the current session without persisting when args are provided', async () => {
     const setValue = vi.fn();
     const switchModel = vi.fn().mockResolvedValue(undefined);
     mockContext = createMockCommandContext({
@@ -178,15 +179,204 @@ describe('modelCommand', () => {
       'qwen-max',
       undefined,
     );
+    expect(setValue).not.toHaveBeenCalled();
+    expect(result).toEqual({
+      type: 'message',
+      messageType: 'info',
+      content: 'Model: qwen-max',
+    });
+  });
+
+  it('should persist the main model only when --default is provided', async () => {
+    const setValue = vi.fn();
+    const switchModel = vi.fn().mockResolvedValue(undefined);
+    mockContext = createMockCommandContext({
+      invocation: {
+        raw: '/model --default gpt-4',
+        name: 'model',
+        args: '--default gpt-4',
+      },
+      services: {
+        config: {
+          getContentGeneratorConfig: vi.fn().mockReturnValue({
+            model: 'gpt-3.5',
+            authType: AuthType.USE_OPENAI,
+          }),
+          getAvailableModelsForAuthType: vi
+            .fn()
+            .mockReturnValue([{ id: 'gpt-4', label: 'GPT-4' }]),
+          switchModel,
+        },
+        settings: createMockSettings(setValue),
+      },
+    });
+
+    const result = await modelCommand.action!(mockContext, '--default gpt-4');
+
+    expect(switchModel).toHaveBeenCalledWith(
+      AuthType.USE_OPENAI,
+      'gpt-4',
+      undefined,
+    );
     expect(setValue).toHaveBeenCalledWith(
       expect.any(String),
       'model.name',
-      'qwen-max',
+      'gpt-4',
     );
     expect(result).toEqual({
       type: 'message',
       messageType: 'info',
-      content: 'Model: qwen-max',
+      content: 'Default model: gpt-4',
+    });
+  });
+
+  it('should persist provider-qualified models when --default is provided', async () => {
+    const setValue = vi.fn();
+    const switchModel = vi.fn().mockResolvedValue(undefined);
+    mockContext = createMockCommandContext({
+      invocation: {
+        raw: `/model --default gpt-4(${AuthType.USE_OPENAI})`,
+        name: 'model',
+        args: `--default gpt-4(${AuthType.USE_OPENAI})`,
+      },
+      services: {
+        config: {
+          getContentGeneratorConfig: vi.fn().mockReturnValue({
+            model: 'qwen-plus',
+            authType: AuthType.QWEN_OAUTH,
+          }),
+          getAvailableModelsForAuthType: vi
+            .fn()
+            .mockReturnValue([{ id: 'gpt-4', label: 'GPT-4' }]),
+          switchModel,
+        },
+        settings: createMockSettings(setValue),
+      },
+    });
+
+    const result = await modelCommand.action!(
+      mockContext,
+      `--default gpt-4(${AuthType.USE_OPENAI})`,
+    );
+
+    expect(switchModel).toHaveBeenCalledWith(
+      AuthType.USE_OPENAI,
+      'gpt-4',
+      undefined,
+    );
+    expect(setValue).toHaveBeenCalledWith(
+      expect.any(String),
+      'security.auth.selectedType',
+      AuthType.USE_OPENAI,
+    );
+    expect(setValue).toHaveBeenCalledWith(
+      expect.any(String),
+      'model.name',
+      'gpt-4',
+    );
+    expect(result).toEqual({
+      type: 'message',
+      messageType: 'info',
+      content: 'Default model: gpt-4',
+    });
+  });
+
+  it('should reject qwen-oauth models when --default is provided', async () => {
+    const setValue = vi.fn();
+    const switchModel = vi.fn().mockResolvedValue(undefined);
+    mockContext = createMockCommandContext({
+      invocation: {
+        raw: '/model --default qwen-max',
+        name: 'model',
+        args: '--default qwen-max',
+      },
+      services: {
+        config: {
+          getContentGeneratorConfig: vi.fn().mockReturnValue({
+            model: 'qwen-plus',
+            authType: AuthType.QWEN_OAUTH,
+          }),
+          getAvailableModelsForAuthType: vi.fn(),
+          switchModel,
+        },
+        settings: createMockSettings(setValue),
+      },
+    });
+
+    const result = await modelCommand.action!(
+      mockContext,
+      '--default qwen-max',
+    );
+
+    expect(switchModel).not.toHaveBeenCalled();
+    expect(setValue).not.toHaveBeenCalled();
+    expect(result).toEqual({
+      type: 'message',
+      messageType: 'error',
+      content:
+        'Qwen OAuth free tier was discontinued on 2026-04-15. Please select a model from another provider.',
+    });
+  });
+
+  it('should return settings error when --default is used without settings', async () => {
+    const switchModel = vi.fn();
+    mockContext = createMockCommandContext({
+      invocation: {
+        raw: '/model --default gpt-4',
+        name: 'model',
+        args: '--default gpt-4',
+      },
+      services: {
+        config: {
+          getContentGeneratorConfig: vi.fn().mockReturnValue({
+            model: 'gpt-3.5',
+            authType: AuthType.USE_OPENAI,
+          }),
+          switchModel,
+        },
+      },
+    });
+    mockContext.services.settings = undefined as unknown as LoadedSettings;
+
+    const result = await modelCommand.action!(mockContext, '--default gpt-4');
+
+    expect(switchModel).not.toHaveBeenCalled();
+    expect(result).toEqual({
+      type: 'message',
+      messageType: 'error',
+      content: 'Settings service not available.',
+    });
+  });
+
+  it('should return usage when --default is missing a model id', async () => {
+    const setValue = vi.fn();
+    const switchModel = vi.fn();
+    mockContext = createMockCommandContext({
+      invocation: {
+        raw: '/model --default',
+        name: 'model',
+        args: '--default',
+      },
+      services: {
+        config: {
+          getContentGeneratorConfig: vi.fn().mockReturnValue({
+            model: 'qwen-plus',
+            authType: AuthType.QWEN_OAUTH,
+          }),
+          switchModel,
+        },
+        settings: createMockSettings(setValue),
+      },
+    });
+
+    const result = await modelCommand.action!(mockContext, '--default');
+
+    expect(switchModel).not.toHaveBeenCalled();
+    expect(setValue).not.toHaveBeenCalled();
+    expect(result).toEqual({
+      type: 'message',
+      messageType: 'error',
+      content: 'Usage: /model --default <model-id>',
     });
   });
 
@@ -343,7 +533,7 @@ describe('modelCommand', () => {
     });
   });
 
-  it('should switch provider-qualified models through switchModel', async () => {
+  it('should switch provider-qualified models for the current session without persisting', async () => {
     const setValue = vi.fn();
     const switchModel = vi.fn().mockResolvedValue(undefined);
     mockContext = createMockCommandContext({
@@ -378,16 +568,7 @@ describe('modelCommand', () => {
       'gpt-4',
       undefined,
     );
-    expect(setValue).toHaveBeenCalledWith(
-      expect.any(String),
-      'security.auth.selectedType',
-      AuthType.USE_OPENAI,
-    );
-    expect(setValue).toHaveBeenCalledWith(
-      expect.any(String),
-      'model.name',
-      'gpt-4',
-    );
+    expect(setValue).not.toHaveBeenCalled();
     expect(result).toEqual({
       type: 'message',
       messageType: 'info',
@@ -572,11 +753,7 @@ describe('modelCommand', () => {
       '--fast-model',
       undefined,
     );
-    expect(setValue).toHaveBeenCalledWith(
-      expect.any(String),
-      'model.name',
-      '--fast-model',
-    );
+    expect(setValue).not.toHaveBeenCalled();
     expect(result).toEqual({
       type: 'message',
       messageType: 'info',
@@ -604,7 +781,11 @@ describe('modelCommand', () => {
       expect(result).toEqual({
         type: 'message',
         messageType: 'info',
-        content: expect.stringContaining('qwen-max'),
+        content:
+          'Current model: qwen-max\n' +
+          'Use "/model <model-id>" to switch models (session only), ' +
+          '"/model --default <model-id>" to persist, or ' +
+          '"/model --fast <model-id>" to set the fast model.',
       });
       expect((result as { type: string }).type).toBe('message');
     });