✨ feat: add support xAI provider (#4627)

* ✨ feat: add support xAI provider

* 💄 style: update model info

* 🐛 fix: fix CI error

* 👷 build: add docker env

* 💄 style: add desc for xAI

Dockerfile

@@ -200,6 +200,8 @@ ENV \
     UPSTAGE_API_KEY="" UPSTAGE_MODEL_LIST="" \
     # Wenxin
     WENXIN_ACCESS_KEY="" WENXIN_SECRET_KEY="" WENXIN_MODEL_LIST="" \
+    # xAI
+    XAI_API_KEY="" XAI_MODEL_LIST="" \
     # 01.AI
     ZEROONE_API_KEY="" ZEROONE_MODEL_LIST="" \
     # Zhipu

Dockerfile.database

@@ -235,6 +235,8 @@ ENV \
     UPSTAGE_API_KEY="" UPSTAGE_MODEL_LIST="" \
     # Wenxin
     WENXIN_ACCESS_KEY="" WENXIN_SECRET_KEY="" WENXIN_MODEL_LIST="" \
+    # xAI
+    XAI_API_KEY="" XAI_MODEL_LIST="" \
     # 01.AI
     ZEROONE_API_KEY="" ZEROONE_MODEL_LIST="" \
     # Zhipu

src/app/(main)/settings/llm/ProviderList/providers.tsx

@@ -23,6 +23,7 @@ import {
   TaichuProviderCard,
   TogetherAIProviderCard,
   UpstageProviderCard,
+  XAIProviderCard,
   ZeroOneProviderCard,
   ZhiPuProviderCard,
 } from '@/config/modelProviders';
@@ -70,6 +71,7 @@ export const useProviderList = (): ProviderItem[] => {
       MistralProviderCard,
       Ai21ProviderCard,
       UpstageProviderCard,
+      XAIProviderCard,
       QwenProviderCard,
       WenxinProvider,
       HunyuanProviderCard,

src/config/llm.ts

@@ -153,6 +153,10 @@ export const getLLMConfig = () => {
       SENSENOVA_ACCESS_KEY_ID: z.string().optional(),
       SENSENOVA_ACCESS_KEY_SECRET: z.string().optional(),
       SENSENOVA_MODEL_LIST: z.string().optional(),
+
+      ENABLED_XAI: z.boolean(),
+      XAI_API_KEY: z.string().optional(),
+      XAI_MODEL_LIST: z.string().optional(),
     },
     runtimeEnv: {
       API_KEY_SELECT_MODE: process.env.API_KEY_SELECT_MODE,
@@ -304,6 +308,10 @@ export const getLLMConfig = () => {
       SENSENOVA_ACCESS_KEY_ID: process.env.SENSENOVA_ACCESS_KEY_ID,
       SENSENOVA_ACCESS_KEY_SECRET: process.env.SENSENOVA_ACCESS_KEY_SECRET,
       SENSENOVA_MODEL_LIST: process.env.SENSENOVA_MODEL_LIST,
+
+      ENABLED_XAI: !!process.env.XAI_API_KEY,
+      XAI_API_KEY: process.env.XAI_API_KEY,
+      XAI_MODEL_LIST: process.env.XAI_MODEL_LIST,
     },
   });
 };

src/config/modelProviders/index.ts

@@ -31,6 +31,7 @@ import TaichuProvider from './taichu';
 import TogetherAIProvider from './togetherai';
 import UpstageProvider from './upstage';
 import WenxinProvider from './wenxin';
+import XAIProvider from './xai';
 import ZeroOneProvider from './zeroone';
 import ZhiPuProvider from './zhipu';
 
@@ -53,6 +54,7 @@ export const LOBE_DEFAULT_MODEL_LIST: ChatModelCard[] = [
   PerplexityProvider.chatModels,
   AnthropicProvider.chatModels,
   HuggingFaceProvider.chatModels,
+  XAIProvider.chatModels,
   ZeroOneProvider.chatModels,
   StepfunProvider.chatModels,
   NovitaProvider.chatModels,
@@ -88,6 +90,7 @@ export const DEFAULT_MODEL_PROVIDER_LIST = [
   MistralProvider,
   Ai21Provider,
   UpstageProvider,
+  XAIProvider,
   QwenProvider,
   WenxinProvider,
   HunyuanProvider,
@@ -145,5 +148,6 @@ export { default as TaichuProviderCard } from './taichu';
 export { default as TogetherAIProviderCard } from './togetherai';
 export { default as UpstageProviderCard } from './upstage';
 export { default as WenxinProviderCard } from './wenxin';
+export { default as XAIProviderCard } from './xai';
 export { default as ZeroOneProviderCard } from './zeroone';
 export { default as ZhiPuProviderCard } from './zhipu';

src/config/modelProviders/xai.ts

@@ -0,0 +1,29 @@
+import { ModelProviderCard } from '@/types/llm';
+
+// ref: https://x.ai/about
+const XAI: ModelProviderCard = {
+  chatModels: [
+    {
+      description: '拥有与 Grok 2 相当的性能，但具有更高的效率、速度和功能。',
+      displayName: 'Grok Beta',
+      enabled: true,
+      functionCall: true,
+      id: 'grok-beta',
+      pricing: {
+        input: 5,
+        output: 15,
+      },
+      tokens: 131_072,
+    },
+  ],
+  checkModel: 'grok-beta',
+  description:
+    'xAI 是一家致力于构建人工智能以加速人类科学发现的公司。我们的使命是推动我们对宇宙的共同理解。',
+  id: 'xai',
+  modelList: { showModelFetcher: true },
+  modelsUrl: 'https://docs.x.ai/docs#models',
+  name: 'xAI',
+  url: 'https://console.x.ai',
+};
+
+export default XAI;

src/const/settings/llm.ts

@@ -29,6 +29,7 @@ import {
   TogetherAIProviderCard,
   UpstageProviderCard,
   WenxinProviderCard,
+  XAIProviderCard,
   ZeroOneProviderCard,
   ZhiPuProviderCard,
   filterEnabledModels,
@@ -161,6 +162,10 @@ export const DEFAULT_LLM_CONFIG: UserModelProviderConfig = {
     enabled: false,
     enabledModels: filterEnabledModels(WenxinProviderCard),
   },
+  xai: {
+    enabled: false,
+    enabledModels: filterEnabledModels(XAIProviderCard),
+  },
   zeroone: {
     enabled: false,
     enabledModels: filterEnabledModels(ZeroOneProviderCard),

src/libs/agent-runtime/AgentRuntime.ts

@@ -42,6 +42,7 @@ import {
   TextToSpeechPayload,
 } from './types';
 import { LobeUpstageAI } from './upstage';
+import { LobeXAI } from './xai';
 import { LobeZeroOneAI } from './zeroone';
 import { LobeZhipuAI } from './zhipu';
 
@@ -156,6 +157,7 @@ class AgentRuntime {
       taichu: Partial<ClientOptions>;
       togetherai: Partial<ClientOptions>;
       upstage: Partial<ClientOptions>;
+      xai: Partial<ClientOptions>;
       zeroone: Partial<ClientOptions>;
       zhipu: Partial<ClientOptions>;
     }>,
@@ -324,6 +326,11 @@ class AgentRuntime {
         break;
       }
 
+      case ModelProvider.XAI: {
+        runtimeModel = new LobeXAI(params.xai);
+        break;
+      }
+
       case ModelProvider.Cloudflare: {
         runtimeModel = new LobeCloudflareAI(params.cloudflare ?? {});
         break;

src/libs/agent-runtime/types/type.ts

@@ -53,6 +53,7 @@ export enum ModelProvider {
   TogetherAI = 'togetherai',
   Upstage = 'upstage',
   Wenxin = 'wenxin',
+  XAI = 'xai',
   ZeroOne = 'zeroone',
   ZhiPu = 'zhipu',
 }

src/libs/agent-runtime/xai/index.test.ts

@@ -0,0 +1,255 @@
+// @vitest-environment node
+import OpenAI from 'openai';
+import { Mock, afterEach, beforeEach, describe, expect, it, vi } from 'vitest';
+
+import {
+  ChatStreamCallbacks,
+  LobeOpenAICompatibleRuntime,
+  ModelProvider,
+} from '@/libs/agent-runtime';
+
+import * as debugStreamModule from '../utils/debugStream';
+import { LobeXAI } from './index';
+
+const provider = ModelProvider.XAI;
+const defaultBaseURL = 'https://api.x.ai/v1';
+
+const bizErrorType = 'ProviderBizError';
+const invalidErrorType = 'InvalidProviderAPIKey';
+
+// Mock the console.error to avoid polluting test output
+vi.spyOn(console, 'error').mockImplementation(() => {});
+
+let instance: LobeOpenAICompatibleRuntime;
+
+beforeEach(() => {
+  instance = new LobeXAI({ apiKey: 'test' });
+
+  // 使用 vi.spyOn 来模拟 chat.completions.create 方法
+  vi.spyOn(instance['client'].chat.completions, 'create').mockResolvedValue(
+    new ReadableStream() as any,
+  );
+});
+
+afterEach(() => {
+  vi.clearAllMocks();
+});
+
+describe('LobeXAI', () => {
+  describe('init', () => {
+    it('should correctly initialize with an API key', async () => {
+      const instance = new LobeXAI({ apiKey: 'test_api_key' });
+      expect(instance).toBeInstanceOf(LobeXAI);
+      expect(instance.baseURL).toEqual(defaultBaseURL);
+    });
+  });
+
+  describe('chat', () => {
+    describe('Error', () => {
+      it('should return OpenAIBizError with an openai error response when OpenAI.APIError is thrown', async () => {
+        // Arrange
+        const apiError = new OpenAI.APIError(
+          400,
+          {
+            status: 400,
+            error: {
+              message: 'Bad Request',
+            },
+          },
+          'Error message',
+          {},
+        );
+
+        vi.spyOn(instance['client'].chat.completions, 'create').mockRejectedValue(apiError);
+
+        // Act
+        try {
+          await instance.chat({
+            messages: [{ content: 'Hello', role: 'user' }],
+            model: 'grok-beta',
+            temperature: 0,
+          });
+        } catch (e) {
+          expect(e).toEqual({
+            endpoint: defaultBaseURL,
+            error: {
+              error: { message: 'Bad Request' },
+              status: 400,
+            },
+            errorType: bizErrorType,
+            provider,
+          });
+        }
+      });
+
+      it('should throw AgentRuntimeError with NoOpenAIAPIKey if no apiKey is provided', async () => {
+        try {
+          new LobeXAI({});
+        } catch (e) {
+          expect(e).toEqual({ errorType: invalidErrorType });
+        }
+      });
+
+      it('should return OpenAIBizError with the cause when OpenAI.APIError is thrown with cause', async () => {
+        // Arrange
+        const errorInfo = {
+          stack: 'abc',
+          cause: {
+            message: 'api is undefined',
+          },
+        };
+        const apiError = new OpenAI.APIError(400, errorInfo, 'module error', {});
+
+        vi.spyOn(instance['client'].chat.completions, 'create').mockRejectedValue(apiError);
+
+        // Act
+        try {
+          await instance.chat({
+            messages: [{ content: 'Hello', role: 'user' }],
+            model: 'grok-beta',
+            temperature: 0,
+          });
+        } catch (e) {
+          expect(e).toEqual({
+            endpoint: defaultBaseURL,
+            error: {
+              cause: { message: 'api is undefined' },
+              stack: 'abc',
+            },
+            errorType: bizErrorType,
+            provider,
+          });
+        }
+      });
+
+      it('should return OpenAIBizError with an cause response with desensitize Url', async () => {
+        // Arrange
+        const errorInfo = {
+          stack: 'abc',
+          cause: { message: 'api is undefined' },
+        };
+        const apiError = new OpenAI.APIError(400, errorInfo, 'module error', {});
+
+        instance = new LobeXAI({
+          apiKey: 'test',
+
+          baseURL: 'https://api.abc.com/v1',
+        });
+
+        vi.spyOn(instance['client'].chat.completions, 'create').mockRejectedValue(apiError);
+
+        // Act
+        try {
+          await instance.chat({
+            messages: [{ content: 'Hello', role: 'user' }],
+            model: 'grok-beta',
+            temperature: 0,
+          });
+        } catch (e) {
+          expect(e).toEqual({
+            endpoint: 'https://api.***.com/v1',
+            error: {
+              cause: { message: 'api is undefined' },
+              stack: 'abc',
+            },
+            errorType: bizErrorType,
+            provider,
+          });
+        }
+      });
+
+      it('should throw an InvalidXAIAPIKey error type on 401 status code', async () => {
+        // Mock the API call to simulate a 401 error
+        const error = new Error('Unauthorized') as any;
+        error.status = 401;
+        vi.mocked(instance['client'].chat.completions.create).mockRejectedValue(error);
+
+        try {
+          await instance.chat({
+            messages: [{ content: 'Hello', role: 'user' }],
+            model: 'grok-beta',
+            temperature: 0,
+          });
+        } catch (e) {
+          // Expect the chat method to throw an error with InvalidXAIAPIKey
+          expect(e).toEqual({
+            endpoint: defaultBaseURL,
+            error: new Error('Unauthorized'),
+            errorType: invalidErrorType,
+            provider,
+          });
+        }
+      });
+
+      it('should return AgentRuntimeError for non-OpenAI errors', async () => {
+        // Arrange
+        const genericError = new Error('Generic Error');
+
+        vi.spyOn(instance['client'].chat.completions, 'create').mockRejectedValue(genericError);
+
+        // Act
+        try {
+          await instance.chat({
+            messages: [{ content: 'Hello', role: 'user' }],
+            model: 'grok-beta',
+            temperature: 0,
+          });
+        } catch (e) {
+          expect(e).toEqual({
+            endpoint: defaultBaseURL,
+            errorType: 'AgentRuntimeError',
+            provider,
+            error: {
+              name: genericError.name,
+              cause: genericError.cause,
+              message: genericError.message,
+              stack: genericError.stack,
+            },
+          });
+        }
+      });
+    });
+
+    describe('DEBUG', () => {
+      it('should call debugStream and return StreamingTextResponse when DEBUG_XAI_CHAT_COMPLETION is 1', async () => {
+        // Arrange
+        const mockProdStream = new ReadableStream() as any; // 模拟的 prod 流
+        const mockDebugStream = new ReadableStream({
+          start(controller) {
+            controller.enqueue('Debug stream content');
+            controller.close();
+          },
+        }) as any;
+        mockDebugStream.toReadableStream = () => mockDebugStream; // 添加 toReadableStream 方法
+
+        // 模拟 chat.completions.create 返回值，包括模拟的 tee 方法
+        (instance['client'].chat.completions.create as Mock).mockResolvedValue({
+          tee: () => [mockProdStream, { toReadableStream: () => mockDebugStream }],
+        });
+
+        // 保存原始环境变量值
+        const originalDebugValue = process.env.DEBUG_XAI_CHAT_COMPLETION;
+
+        // 模拟环境变量
+        process.env.DEBUG_XAI_CHAT_COMPLETION = '1';
+        vi.spyOn(debugStreamModule, 'debugStream').mockImplementation(() => Promise.resolve());
+
+        // 执行测试
+        // 运行你的测试函数，确保它会在条件满足时调用 debugStream
+        // 假设的测试函数调用，你可能需要根据实际情况调整
+        await instance.chat({
+          messages: [{ content: 'Hello', role: 'user' }],
+          model: 'grok-beta',
+          stream: true,
+          temperature: 0,
+        });
+
+        // 验证 debugStream 被调用
+        expect(debugStreamModule.debugStream).toHaveBeenCalled();
+
+        // 恢复原始环境变量值
+        process.env.DEBUG_XAI_CHAT_COMPLETION = originalDebugValue;
+      });
+    });
+  });
+});

src/libs/agent-runtime/xai/index.ts

@@ -0,0 +1,10 @@
+import { ModelProvider } from '../types';
+import { LobeOpenAICompatibleFactory } from '../utils/openaiCompatibleFactory';
+
+export const LobeXAI = LobeOpenAICompatibleFactory({
+  baseURL: 'https://api.x.ai/v1',
+  debug: {
+    chatCompletion: () => process.env.DEBUG_XAI_CHAT_COMPLETION === '1',
+  },
+  provider: ModelProvider.XAI,
+});

src/server/globalConfig/index.ts

@@ -33,6 +33,7 @@ import {
   TogetherAIProviderCard,
   UpstageProviderCard,
   WenxinProviderCard,
+  XAIProviderCard,
   ZeroOneProviderCard,
   ZhiPuProviderCard,
 } from '@/config/modelProviders';
@@ -146,6 +147,9 @@ export const getServerGlobalConfig = () => {
 
     ENABLED_HUGGINGFACE,
     HUGGINGFACE_MODEL_LIST,
+
+    ENABLED_XAI,
+    XAI_MODEL_LIST,
   } = getLLMConfig();
 
   const config: GlobalServerConfig = {
@@ -399,6 +403,14 @@ export const getServerGlobalConfig = () => {
           modelString: WENXIN_MODEL_LIST,
         }),
       },
+      xai: {
+        enabled: ENABLED_XAI,
+        enabledModels: extractEnabledModels(XAI_MODEL_LIST),
+        serverModelCards: transformToChatModelCards({
+          defaultChatModels: XAIProviderCard.chatModels,
+          modelString: XAI_MODEL_LIST,
+        }),
+      },
       zeroone: {
         enabled: ENABLED_ZEROONE,
         enabledModels: extractEnabledModels(ZEROONE_MODEL_LIST),

src/server/modules/AgentRuntime/index.ts

@@ -286,6 +286,13 @@ const getLlmOptionsFromPayload = (provider: string, payload: JWTPayload) => {
 
       const apiKey = sensenovaAccessKeyID + ':' + sensenovaAccessKeySecret;
 
+      return { apiKey };
+    }
+    case ModelProvider.XAI: {
+      const { XAI_API_KEY } = getLLMConfig();
+
+      const apiKey = apiKeyManager.pick(payload?.apiKey || XAI_API_KEY);
+
       return { apiKey };
     }
   }

src/types/user/settings/keyVaults.ts

@@ -65,6 +65,7 @@ export interface UserKeyVaults {
   togetherai?: OpenAICompatibleKeyVault;
   upstage?: OpenAICompatibleKeyVault;
   wenxin?: WenxinKeyVault;
+  xai?: OpenAICompatibleKeyVault;
   zeroone?: OpenAICompatibleKeyVault;
   zhipu?: OpenAICompatibleKeyVault;
 }