diff --git a/packages/cli/src/utils/modelConfigUtils.test.ts b/packages/cli/src/utils/modelConfigUtils.test.ts
index 97cea9974..abd556308 100644
--- a/packages/cli/src/utils/modelConfigUtils.test.ts
+++ b/packages/cli/src/utils/modelConfigUtils.test.ts
@@ -442,6 +442,187 @@ describe('modelConfigUtils', () => {
       );
     });
 
+    it('should find modelProvider from OPENAI_MODEL when argv.model is not provided', () => {
+      const argv = {};
+      const modelProvider: ProviderModelConfig = {
+        id: 'env-openai-model',
+        name: 'Env OpenAI Model',
+        generationConfig: {
+          samplingParams: { temperature: 0.6 },
+        },
+      };
+      const settings = makeMockSettings({
+        model: { name: 'settings-model' },
+        modelProviders: {
+          [AuthType.USE_OPENAI]: [
+            { id: 'settings-model', name: 'Settings Model' },
+            modelProvider,
+          ],
+        },
+      });
+      const selectedAuthType = AuthType.USE_OPENAI;
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: {
+          model: 'env-openai-model',
+          apiKey: '',
+          baseUrl: '',
+        },
+        sources: {},
+        warnings: [],
+      });
+
+      resolveCliGenerationConfig({
+        argv,
+        settings,
+        selectedAuthType,
+        env: { OPENAI_MODEL: 'env-openai-model' },
+      });
+
+      expect(vi.mocked(resolveModelConfig)).toHaveBeenCalledWith(
+        expect.objectContaining({
+          modelProvider,
+        }),
+      );
+    });
+
+    it('should find modelProvider from QWEN_MODEL when OPENAI_MODEL is not provided', () => {
+      const argv = {};
+      const modelProvider: ProviderModelConfig = {
+        id: 'qwen-env-model',
+        name: 'Qwen Env Model',
+        generationConfig: {
+          samplingParams: { temperature: 0.7 },
+        },
+      };
+      const settings = makeMockSettings({
+        model: { name: 'settings-model' },
+        modelProviders: {
+          [AuthType.USE_OPENAI]: [
+            { id: 'settings-model', name: 'Settings Model' },
+            modelProvider,
+          ],
+        },
+      });
+      const selectedAuthType = AuthType.USE_OPENAI;
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: {
+          model: 'qwen-env-model',
+          apiKey: '',
+          baseUrl: '',
+        },
+        sources: {},
+        warnings: [],
+      });
+
+      resolveCliGenerationConfig({
+        argv,
+        settings,
+        selectedAuthType,
+        env: { QWEN_MODEL: 'qwen-env-model' },
+      });
+
+      expect(vi.mocked(resolveModelConfig)).toHaveBeenCalledWith(
+        expect.objectContaining({
+          modelProvider,
+        }),
+      );
+    });
+
+    it('should prefer OPENAI_MODEL over QWEN_MODEL and settings.model.name for USE_OPENAI provider lookup', () => {
+      const argv = {};
+      const openAIProvider: ProviderModelConfig = {
+        id: 'openai-env-model',
+        name: 'OpenAI Env Model',
+      };
+      const qwenProvider: ProviderModelConfig = {
+        id: 'qwen-env-model',
+        name: 'Qwen Env Model',
+      };
+      const settings = makeMockSettings({
+        model: { name: 'settings-model' },
+        modelProviders: {
+          [AuthType.USE_OPENAI]: [
+            { id: 'settings-model', name: 'Settings Model' },
+            qwenProvider,
+            openAIProvider,
+          ],
+        },
+      });
+      const selectedAuthType = AuthType.USE_OPENAI;
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: {
+          model: 'openai-env-model',
+          apiKey: '',
+          baseUrl: '',
+        },
+        sources: {},
+        warnings: [],
+      });
+
+      resolveCliGenerationConfig({
+        argv,
+        settings,
+        selectedAuthType,
+        env: {
+          OPENAI_MODEL: 'openai-env-model',
+          QWEN_MODEL: 'qwen-env-model',
+        },
+      });
+
+      expect(vi.mocked(resolveModelConfig)).toHaveBeenCalledWith(
+        expect.objectContaining({
+          modelProvider: openAIProvider,
+        }),
+      );
+    });
+
+    it('should ignore OPENAI_MODEL for non-USE_OPENAI provider lookup', () => {
+      const argv = {};
+      const settingsModelProvider: ProviderModelConfig = {
+        id: 'settings-model',
+        name: 'Settings Model',
+      };
+      const unrelatedOpenAIProvider: ProviderModelConfig = {
+        id: 'openai-env-model',
+        name: 'OpenAI Env Model',
+      };
+      const settings = makeMockSettings({
+        model: { name: 'settings-model' },
+        modelProviders: {
+          [AuthType.USE_ANTHROPIC]: [
+            settingsModelProvider,
+            unrelatedOpenAIProvider,
+          ],
+        },
+      });
+      const selectedAuthType = AuthType.USE_ANTHROPIC;
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: {
+          model: 'settings-model',
+          apiKey: '',
+          baseUrl: '',
+        },
+        sources: {},
+        warnings: [],
+      });
+
+      resolveCliGenerationConfig({
+        argv,
+        settings,
+        selectedAuthType,
+        env: { OPENAI_MODEL: 'openai-env-model' },
+      });
+
+      expect(vi.mocked(resolveModelConfig)).toHaveBeenCalledWith(
+        expect.objectContaining({
+          modelProvider: settingsModelProvider,
+        }),
+      );
+    });
     it('should not find modelProvider when authType is undefined', () => {
       const argv = { model: 'test-model' };
       const settings = makeMockSettings({
@@ -723,5 +904,71 @@ describe('modelConfigUtils', () => {
         }),
       );
     });
+
+    it('should respect precedence: argv.model > OPENAI_MODEL > QWEN_MODEL > settings.model.name', () => {
+      const mockSettings = makeMockSettings({
+        model: { name: 'settings-model' },
+        modelProviders: {
+          [AuthType.USE_OPENAI]: [
+            { id: 'settings-model' } as ProviderModelConfig,
+            { id: 'openai-env-model' } as ProviderModelConfig,
+            { id: 'qwen-env-model' } as ProviderModelConfig,
+            { id: 'cli-model' } as ProviderModelConfig,
+          ],
+        },
+      });
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: { model: 'cli-model', apiKey: '', baseUrl: '' },
+        sources: {},
+        warnings: [],
+      });
+      const result1 = resolveCliGenerationConfig({
+        argv: { model: 'cli-model' },
+        settings: mockSettings,
+        selectedAuthType: AuthType.USE_OPENAI,
+        env: { OPENAI_MODEL: 'openai-env-model' },
+      });
+      expect(result1.model).toBe('cli-model');
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: { model: 'openai-env-model', apiKey: '', baseUrl: '' },
+        sources: {},
+        warnings: [],
+      });
+      const result2 = resolveCliGenerationConfig({
+        argv: {},
+        settings: mockSettings,
+        selectedAuthType: AuthType.USE_OPENAI,
+        env: { OPENAI_MODEL: 'openai-env-model', QWEN_MODEL: 'qwen-env-model' },
+      });
+      expect(result2.model).toBe('openai-env-model');
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: { model: 'openai-env-model', apiKey: '', baseUrl: '' },
+        sources: {},
+        warnings: [],
+      });
+      const result3 = resolveCliGenerationConfig({
+        argv: {},
+        settings: mockSettings,
+        selectedAuthType: AuthType.USE_OPENAI,
+        env: { OPENAI_MODEL: 'openai-env-model' },
+      });
+      expect(result3.model).toBe('openai-env-model');
+
+      vi.mocked(resolveModelConfig).mockReturnValue({
+        config: { model: 'settings-model', apiKey: '', baseUrl: '' },
+        sources: {},
+        warnings: [],
+      });
+      const result4 = resolveCliGenerationConfig({
+        argv: {},
+        settings: mockSettings,
+        selectedAuthType: AuthType.USE_OPENAI,
+        env: {},
+      });
+      expect(result4.model).toBe('settings-model');
+    });
   });
 });
diff --git a/packages/cli/src/utils/modelConfigUtils.ts b/packages/cli/src/utils/modelConfigUtils.ts
index aa5ac5e82..83dbae1cc 100644
--- a/packages/cli/src/utils/modelConfigUtils.ts
+++ b/packages/cli/src/utils/modelConfigUtils.ts
@@ -100,8 +100,13 @@ export function resolveCliGenerationConfig(
   if (authType && settings.modelProviders) {
     const providers = settings.modelProviders[authType];
     if (providers && Array.isArray(providers)) {
-      // Try to find by requested model (from CLI or settings)
-      const requestedModel = argv.model || settings.model?.name;
+      const requestedModel =
+        authType === AuthType.USE_OPENAI
+          ? argv.model ||
+            env['OPENAI_MODEL'] ||
+            env['QWEN_MODEL'] ||
+            settings.model?.name
+          : argv.model || settings.model?.name;
       if (requestedModel) {
         modelProvider = providers.find((p) => p.id === requestedModel) as
           | ProviderModelConfig