@lobehub/chat 1.21.12 → 1.21.14

package/docs/usage/providers/wenxin.zh-CN.mdx

@@ -0,0 +1,73 @@
+---
+title: 在 LobeChat 中使用文心千帆
+description: 学习如何在 LobeChat 中配置和使用文心千帆的API Key，以便开始对话和交互。
+tags:
+  - LobeChat
+  - 百度
+  - 文心千帆
+  - API密钥
+  - Web UI
+---
+
+# 在 LobeChat 中使用文心千帆
+
+<Image
+  cover
+  src={
+    'https://github.com/user-attachments/assets/e43dacf6-313e-499c-8888-f1065c53e424'
+  }
+/>
+
+[文心千帆](https://qianfan.cloud.baidu.com/)是百度推出的一个人工智能大语言模型平台，支持多种应用场景，包括文学创作、商业文案生成、数理逻辑推算等。该平台具备跨模态、跨语言的深度语义理解与生成能力，广泛应用于搜索问答、内容创作和智能办公等领域。
+
+本文将指导你如何在 LobeChat 中使用文心千帆。
+
+<Steps>
+### 步骤一：获得文心千帆的 API Key
+
+- 注册并登录 [百度智能云控制台](https://console.bce.baidu.com/)
+- 进入 `百度智能云千帆 ModelBuilder`
+- 在左侧菜单中选择`应用接入`
+- 创建一个应用
+
+<Image
+  alt={'创建应用'}
+  inStep
+  src={'https://github.com/user-attachments/assets/927b1040-e23f-4919-92e2-80a400db8327'}
+/>
+
+- 创建成功后获取 `API Key` 和 `Secret Key`，并妥善保存
+
+<Image
+  alt={'保存密钥'}
+  inStep
+  src={'https://github.com/user-attachments/assets/242c8134-8de0-4a02-b302-6bd8b19ced3e'}
+/>
+
+### 步骤二：在 LobeChat 中配置文心千帆
+
+- 访问 LobeChat 的`设置`界面
+- 在`语言模型`下找到 `文心千帆` 的设置项
+
+<Image
+  alt={'填入 API 密钥'}
+  inStep
+  src={'https://github.com/user-attachments/assets/e3995de7-38d9-489b-80a2-434477018469'}
+/>
+
+- 填入获得的 `API Key` 和 `Secret Key`
+- 为你的 AI 助手选择一个文心千帆的模型即可开始对话
+
+<Image
+  alt={'选择文心千帆模型并开始对话'}
+  inStep
+  src={'https://github.com/user-attachments/assets/b6e6a3eb-13c6-46f0-9c7c-69a20deae30f'}
+/>
+
+<Callout type={'warning'}>
+  在使用过程中你可能需要向 API 服务提供商付费，请参考文心千帆的相关费用政策。
+</Callout>
+
+</Steps>
+
+至此你已经可以在 LobeChat 中使用文心千帆提供的模型进行对话了。

package/docs/usage/providers/{01ai.mdx → zeroone.mdx}

@@ -1,39 +1,38 @@
 ---
-title: Using Zero One AI API Key in LobeChat
+title: Using 01 AI API Key in LobeChat
 description: >-
-  Learn how to integrate and use Zero One AI in LobeChat with step-by-step
-  instructions. Obtain an API key, configure Zero One AI, and start
+  Learn how to integrate and use 01 AI in LobeChat with step-by-step
+  instructions. Obtain an API key, configure 01 AI, and start
   conversations with AI models.
 tags:
   - 01.AI
-  - Zero One AI
   - Web UI
   - API key
   - AI models
 ---
 
-# Using Zero One AI in LobeChat
+# Using 01 AI in LobeChat
 
 <Image
-  alt={'Using Zero One AI in LobeChat'}
+  alt={'Using 01 AI in LobeChat'}
   cover
   src={'https://github.com/lobehub/lobe-chat/assets/34400653/4485fbc3-c309-4c4e-83ee-cb82392307a1'}
 />
 
-[Zero One AI](https://www.01.ai/) is a global company dedicated to AI 2.0 large model technology and applications. Its billion-parameter Yi-Large closed-source model, when evaluated on Stanford University's English ranking AlpacaEval 2.0, is on par with GPT-4.
+[01 AI](https://www.01.ai/) is a global company dedicated to AI 2.0 large model technology and applications. Its billion-parameter Yi-Large closed-source model, when evaluated on Stanford University's English ranking AlpacaEval 2.0, is on par with GPT-4.
 
-This document will guide you on how to use Zero One AI in LobeChat:
+This document will guide you on how to use 01 AI in LobeChat:
 
 <Steps>
 
-### Step 1: Obtain Zero One AI API Key
+### Step 1: Obtain 01 AI API Key
 
-- Register and log in to the [Zero One AI Large Model Open Platform](https://platform.lingyiwanwu.com/)
+- Register and log in to the [01 AI Large Model Open Platform](https://platform.lingyiwanwu.com/)
 - Go to the `Dashboard` and access the `API Key Management` menu
 - A system-generated API key has been created for you automatically, or you can create a new one on this interface
 
 <Image
-  alt={'Create Zero One AI API Key'}
+  alt={'Create 01 AI API Key'}
   inStep
   src={'https://github.com/lobehub/lobe-chat/assets/34400653/72f165f4-d529-4f01-a3ac-163c66e5ea73'}
 />
@@ -55,10 +54,10 @@ This document will guide you on how to use Zero One AI in LobeChat:
   src={'https://github.com/lobehub/lobe-chat/assets/34400653/f892fe64-c734-4944-91ff-9916a41bd1c9'}
 />
 
-### Step 2: Configure Zero One AI in LobeChat
+### Step 2: Configure 01 AI in LobeChat
 
 - Access the `Settings` interface in LobeChat
-- Find the setting for `Zero One AI` under `Language Model`
+- Find the setting for `01 AI` under `Language Model`
 
 <Image
   alt={'Enter API Key'}
@@ -66,7 +65,7 @@ This document will guide you on how to use Zero One AI in LobeChat:
   src={'https://github.com/lobehub/lobe-chat/assets/34400653/f539d104-6d64-4cc7-8781-3b36b00d32d0'}
 />
 
-- Open Zero One AI and enter the obtained API key
+- Open 01 AI and enter the obtained API key
 - Choose a 01.AI model for your AI assistant to start the conversation
 
 <Image
@@ -76,10 +75,10 @@ This document will guide you on how to use Zero One AI in LobeChat:
 />
 
 <Callout type={'warning'}>
-  During usage, you may need to pay the API service provider. Please refer to Zero One AI's relevant
+  During usage, you may need to pay the API service provider. Please refer to 01 AI's relevant
   fee policies.
 </Callout>
 
 </Steps>
 
-You can now use the models provided by Zero One AI for conversations in LobeChat.
+You can now use the models provided by 01 AI for conversations in LobeChat.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lobehub/chat",
-  "version": "1.21.12",
+  "version": "1.21.14",
   "description": "Lobe Chat - an open-source, high-performance chatbot framework that supports speech synthesis, multimodal, and extensible Function Call plugin system. Supports one-click free deployment of your private ChatGPT/LLM web application.",
   "keywords": [
     "framework",

package/src/app/(main)/chat/(workspace)/@portal/Artifacts/Body/Renderer/index.tsx

@@ -1,3 +1,4 @@
+import { Markdown } from '@lobehub/ui';
 import dynamic from 'next/dynamic';
 import { memo } from 'react';
 
@@ -16,6 +17,10 @@ const Renderer = memo<{ content: string; type?: string }>(({ content, type }) =>
       return <SVGRender content={content} />;
     }
 
+    case 'text/markdown': {
+      return <Markdown>{content}</Markdown>;
+    }
+
     default: {
       return <HTMLRenderer htmlContent={content} />;
     }

package/src/libs/agent-runtime/AgentRuntime.ts

@@ -174,7 +174,7 @@ class AgentRuntime {
       }
 
       case ModelProvider.ZhiPu: {
-        runtimeModel = await LobeZhipuAI.fromAPIKey(params.zhipu);
+        runtimeModel = new LobeZhipuAI(params.zhipu);
         break;
       }
 

package/src/libs/agent-runtime/google/index.ts

@@ -27,7 +27,7 @@ import { ModelProvider } from '../types/type';
 import { AgentRuntimeError } from '../utils/createError';
 import { debugStream } from '../utils/debugStream';
 import { StreamingResponse } from '../utils/response';
-import { GoogleGenerativeAIStream, googleGenAIResultToStream } from '../utils/streams';
+import { GoogleGenerativeAIStream, convertIterableToStream } from '../utils/streams';
 import { parseDataUri } from '../utils/uriParser';
 
 enum HarmCategory {
@@ -97,7 +97,7 @@ export class LobeGoogleAI implements LobeRuntimeAI {
           tools: this.buildGoogleTools(payload.tools),
         });
 
-      const googleStream = googleGenAIResultToStream(geminiStreamResult);
+      const googleStream = convertIterableToStream(geminiStreamResult.stream);
       const [prod, useForDebug] = googleStream.tee();
 
       if (process.env.DEBUG_GOOGLE_CHAT_COMPLETION === '1') {

package/src/libs/agent-runtime/utils/streams/anthropic.ts

@@ -1,6 +1,5 @@
 import Anthropic from '@anthropic-ai/sdk';
 import type { Stream } from '@anthropic-ai/sdk/streaming';
-import { readableFromAsyncIterable } from 'ai';
 
 import { ChatStreamCallbacks } from '../../types';
 import {
@@ -8,6 +7,7 @@ import {
   StreamProtocolToolCallChunk,
   StreamStack,
   StreamToolCallChunkData,
+  convertIterableToStream,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
 } from './protocol';
@@ -96,12 +96,6 @@ export const transformAnthropicStream = (
   }
 };
 
-const chatStreamable = async function* (stream: AsyncIterable<Anthropic.MessageStreamEvent>) {
-  for await (const response of stream) {
-    yield response;
-  }
-};
-
 export const AnthropicStream = (
   stream: Stream<Anthropic.MessageStreamEvent> | ReadableStream,
   callbacks?: ChatStreamCallbacks,
@@ -109,7 +103,7 @@ export const AnthropicStream = (
   const streamStack: StreamStack = { id: '' };
 
   const readableStream =
-    stream instanceof ReadableStream ? stream : readableFromAsyncIterable(chatStreamable(stream));
+    stream instanceof ReadableStream ? stream : convertIterableToStream(stream);
 
   return readableStream
     .pipeThrough(createSSEProtocolTransformer(transformAnthropicStream, streamStack))

package/src/libs/agent-runtime/utils/streams/azureOpenai.ts

@@ -1,5 +1,4 @@
 import { ChatCompletions, ChatCompletionsFunctionToolCall } from '@azure/openai';
-import { readableFromAsyncIterable } from 'ai';
 import OpenAI from 'openai';
 import type { Stream } from 'openai/streaming';
 
@@ -9,6 +8,7 @@ import {
   StreamProtocolToolCallChunk,
   StreamStack,
   StreamToolCallChunkData,
+  convertIterableToStream,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
 } from './protocol';
@@ -69,19 +69,13 @@ const transformOpenAIStream = (chunk: ChatCompletions, stack: StreamStack): Stre
   };
 };
 
-const chatStreamable = async function* (stream: AsyncIterable<OpenAI.ChatCompletionChunk>) {
-  for await (const response of stream) {
-    yield response;
-  }
-};
-
 export const AzureOpenAIStream = (
   stream: Stream<OpenAI.ChatCompletionChunk> | ReadableStream,
   callbacks?: ChatStreamCallbacks,
 ) => {
   const stack: StreamStack = { id: '' };
   const readableStream =
-    stream instanceof ReadableStream ? stream : readableFromAsyncIterable(chatStreamable(stream));
+    stream instanceof ReadableStream ? stream : convertIterableToStream(stream);
 
   return readableStream
     .pipeThrough(createSSEProtocolTransformer(transformOpenAIStream, stack))

package/src/libs/agent-runtime/utils/streams/google-ai.ts

@@ -1,8 +1,4 @@
-import {
-  EnhancedGenerateContentResponse,
-  GenerateContentStreamResult,
-} from '@google/generative-ai';
-import { readableFromAsyncIterable } from 'ai';
+import { EnhancedGenerateContentResponse } from '@google/generative-ai';
 
 import { nanoid } from '@/utils/uuid';
 
@@ -11,7 +7,6 @@ import {
   StreamProtocolChunk,
   StreamStack,
   StreamToolCallChunkData,
-  chatStreamable,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
   generateToolCallId,
@@ -50,12 +45,6 @@ const transformGoogleGenerativeAIStream = (
   };
 };
 
-// only use for debug
-export const googleGenAIResultToStream = (stream: GenerateContentStreamResult) => {
-  // make the response to the streamable format
-  return readableFromAsyncIterable(chatStreamable(stream.stream));
-};
-
 export const GoogleGenerativeAIStream = (
   rawStream: ReadableStream<EnhancedGenerateContentResponse>,
   callbacks?: ChatStreamCallbacks,

package/src/libs/agent-runtime/utils/streams/ollama.ts

@@ -1,4 +1,3 @@
-import { readableFromAsyncIterable } from 'ai';
 import { ChatResponse } from 'ollama/browser';
 
 import { ChatStreamCallbacks } from '@/libs/agent-runtime';
@@ -7,6 +6,7 @@ import { nanoid } from '@/utils/uuid';
 import {
   StreamProtocolChunk,
   StreamStack,
+  convertIterableToStream,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
 } from './protocol';
@@ -20,19 +20,13 @@ const transformOllamaStream = (chunk: ChatResponse, stack: StreamStack): StreamP
   return { data: chunk.message.content, id: stack.id, type: 'text' };
 };
 
-const chatStreamable = async function* (stream: AsyncIterable<ChatResponse>) {
-  for await (const response of stream) {
-    yield response;
-  }
-};
-
 export const OllamaStream = (
   res: AsyncIterable<ChatResponse>,
   cb?: ChatStreamCallbacks,
 ): ReadableStream<string> => {
   const streamStack: StreamStack = { id: 'chat_' + nanoid() };
 
-  return readableFromAsyncIterable(chatStreamable(res))
+  return convertIterableToStream(res)
     .pipeThrough(createSSEProtocolTransformer(transformOllamaStream, streamStack))
     .pipeThrough(createCallbacksTransformer(cb));
 };

package/src/libs/agent-runtime/utils/streams/openai.ts

@@ -1,4 +1,3 @@
-import { readableFromAsyncIterable } from 'ai';
 import OpenAI from 'openai';
 import type { Stream } from 'openai/streaming';
 
@@ -10,6 +9,7 @@ import {
   StreamProtocolToolCallChunk,
   StreamStack,
   StreamToolCallChunkData,
+  convertIterableToStream,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
   generateToolCallId,
@@ -105,12 +105,6 @@ export const transformOpenAIStream = (
   }
 };
 
-const chatStreamable = async function* (stream: AsyncIterable<OpenAI.ChatCompletionChunk>) {
-  for await (const response of stream) {
-    yield response;
-  }
-};
-
 export const OpenAIStream = (
   stream: Stream<OpenAI.ChatCompletionChunk> | ReadableStream,
   callbacks?: ChatStreamCallbacks,
@@ -118,7 +112,7 @@ export const OpenAIStream = (
   const streamStack: StreamStack = { id: '' };
 
   const readableStream =
-    stream instanceof ReadableStream ? stream : readableFromAsyncIterable(chatStreamable(stream));
+    stream instanceof ReadableStream ? stream : convertIterableToStream(stream);
 
   return readableStream
     .pipeThrough(createSSEProtocolTransformer(transformOpenAIStream, streamStack))

package/src/libs/agent-runtime/utils/streams/protocol.ts

@@ -1,3 +1,5 @@
+import { readableFromAsyncIterable } from 'ai';
+
 import { ChatStreamCallbacks } from '@/libs/agent-runtime';
 
 export interface StreamStack {
@@ -42,6 +44,11 @@ export const chatStreamable = async function* <T>(stream: AsyncIterable<T>) {
   }
 };
 
+// make the response to the streamable format
+export const convertIterableToStream = <T>(stream: AsyncIterable<T>) => {
+  return readableFromAsyncIterable(chatStreamable(stream));
+};
+
 export const createSSEProtocolTransformer = (
   transformer: (chunk: any, stack: StreamStack) => StreamProtocolChunk,
   streamStack?: StreamStack,

package/src/libs/agent-runtime/utils/streams/qwen.ts

@@ -1,4 +1,3 @@
-import { readableFromAsyncIterable } from 'ai';
 import { ChatCompletionContentPartText } from 'ai/prompts';
 import OpenAI from 'openai';
 import { ChatCompletionContentPart } from 'openai/resources/index.mjs';
@@ -9,7 +8,7 @@ import {
   StreamProtocolChunk,
   StreamProtocolToolCallChunk,
   StreamToolCallChunkData,
-  chatStreamable,
+  convertIterableToStream,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
   generateToolCallId,
@@ -86,7 +85,7 @@ export const QwenAIStream = (
   callbacks?: ChatStreamCallbacks,
 ) => {
   const readableStream =
-    stream instanceof ReadableStream ? stream : readableFromAsyncIterable(chatStreamable(stream));
+    stream instanceof ReadableStream ? stream : convertIterableToStream(stream);
 
   return readableStream
     .pipeThrough(createSSEProtocolTransformer(transformQwenStream))

package/src/libs/agent-runtime/utils/streams/wenxin.test.ts

@@ -2,8 +2,9 @@ import { describe, expect, it, vi } from 'vitest';
 
 import * as uuidModule from '@/utils/uuid';
 
+import { convertIterableToStream } from '../../utils/streams/protocol';
 import { ChatResp } from '../../wenxin/type';
-import { WenxinResultToStream, WenxinStream } from './wenxin';
+import { WenxinStream } from './wenxin';
 
 const dataStream = [
   {
@@ -95,7 +96,7 @@ describe('WenxinStream', () => {
       },
     };
 
-    const stream = WenxinResultToStream(mockWenxinStream);
+    const stream = convertIterableToStream(mockWenxinStream);
 
     const onStartMock = vi.fn();
     const onTextMock = vi.fn();
@@ -142,7 +143,10 @@ describe('WenxinStream', () => {
 
     expect(onStartMock).toHaveBeenCalledTimes(1);
     expect(onTextMock).toHaveBeenNthCalledWith(1, '"当然可以，"');
-    expect(onTextMock).toHaveBeenNthCalledWith(2, '"以下是一些建议的自驾游路线，它们涵盖了各种不同的风景和文化体验：\\n\\n1. **西安-敦煌历史文化之旅**：\\n\\n\\n\\t* 路线：西安"');
+    expect(onTextMock).toHaveBeenNthCalledWith(
+      2,
+      '"以下是一些建议的自驾游路线，它们涵盖了各种不同的风景和文化体验：\\n\\n1. **西安-敦煌历史文化之旅**：\\n\\n\\n\\t* 路线：西安"',
+    );
     expect(onTokenMock).toHaveBeenCalledTimes(6);
     expect(onCompletionMock).toHaveBeenCalledTimes(1);
   });

package/src/libs/agent-runtime/utils/streams/wenxin.ts

@@ -1,5 +1,3 @@
-import { readableFromAsyncIterable } from 'ai';
-
 import { ChatStreamCallbacks } from '@/libs/agent-runtime';
 import { nanoid } from '@/utils/uuid';
 
@@ -7,7 +5,6 @@ import { ChatResp } from '../../wenxin/type';
 import {
   StreamProtocolChunk,
   StreamStack,
-  chatStreamable,
   createCallbacksTransformer,
   createSSEProtocolTransformer,
 } from './protocol';
@@ -29,11 +26,6 @@ const transformERNIEBotStream = (chunk: ChatResp): StreamProtocolChunk => {
   };
 };
 
-export const WenxinResultToStream = (stream: AsyncIterable<ChatResp>) => {
-  // make the response to the streamable format
-  return readableFromAsyncIterable(chatStreamable(stream));
-};
-
 export const WenxinStream = (
   rawStream: ReadableStream<ChatResp>,
   callbacks?: ChatStreamCallbacks,

package/src/libs/agent-runtime/wenxin/index.ts

@@ -10,7 +10,8 @@ import { ChatCompetitionOptions, ChatStreamPayload } from '../types';
 import { AgentRuntimeError } from '../utils/createError';
 import { debugStream } from '../utils/debugStream';
 import { StreamingResponse } from '../utils/response';
-import { WenxinResultToStream, WenxinStream } from '../utils/streams/wenxin';
+import { convertIterableToStream } from '../utils/streams';
+import { WenxinStream } from '../utils/streams/wenxin';
 import { ChatResp } from './type';
 
 interface ChatErrorCode {
@@ -46,7 +47,7 @@ export class LobeWenxinAI implements LobeRuntimeAI {
         payload.model,
       );
 
-      const wenxinStream = WenxinResultToStream(result as AsyncIterable<ChatResp>);
+      const wenxinStream = convertIterableToStream(result as AsyncIterable<ChatResp>);
 
       const [prod, useForDebug] = wenxinStream.tee();
 

package/src/libs/agent-runtime/zhipu/index.test.ts

@@ -2,7 +2,7 @@
 import { OpenAI } from 'openai';
 import { Mock, afterEach, beforeEach, describe, expect, it, vi } from 'vitest';
 
-import { ChatStreamCallbacks, LobeOpenAI } from '@/libs/agent-runtime';
+import { ChatStreamCallbacks, LobeOpenAI, LobeOpenAICompatibleRuntime } from '@/libs/agent-runtime';
 import * as debugStreamModule from '@/libs/agent-runtime/utils/debugStream';
 
 import * as authTokenModule from './authToken';
@@ -24,28 +24,11 @@ describe('LobeZhipuAI', () => {
     vi.restoreAllMocks();
   });
 
-  describe('fromAPIKey', () => {
-    it('should correctly initialize with an API key', async () => {
-      const lobeZhipuAI = await LobeZhipuAI.fromAPIKey({ apiKey: 'test_api_key' });
-      expect(lobeZhipuAI).toBeInstanceOf(LobeZhipuAI);
-      expect(lobeZhipuAI.baseURL).toEqual('https://open.bigmodel.cn/api/paas/v4');
-    });
-
-    it('should throw an error if API key is invalid', async () => {
-      vi.spyOn(authTokenModule, 'generateApiToken').mockRejectedValue(new Error('Invalid API Key'));
-      try {
-        await LobeZhipuAI.fromAPIKey({ apiKey: 'asd' });
-      } catch (e) {
-        expect(e).toEqual({ errorType: invalidErrorType });
-      }
-    });
-  });
-
   describe('chat', () => {
-    let instance: LobeZhipuAI;
+    let instance: LobeOpenAICompatibleRuntime;
 
     beforeEach(async () => {
-      instance = await LobeZhipuAI.fromAPIKey({
+      instance = new LobeZhipuAI({
         apiKey: 'test_api_key',
       });
 
@@ -131,9 +114,9 @@ describe('LobeZhipuAI', () => {
       const calledWithParams = spyOn.mock.calls[0][0];
 
       expect(calledWithParams.messages[1].content).toEqual([{ type: 'text', text: 'Hello again' }]);
-      expect(calledWithParams.temperature).toBeUndefined(); // temperature 0 should be undefined
+      expect(calledWithParams.temperature).toBe(0); // temperature 0 should be undefined
       expect((calledWithParams as any).do_sample).toBeTruthy(); // temperature 0 should be undefined
-      expect(calledWithParams.top_p).toEqual(0.99); // top_p should be transformed correctly
+      expect(calledWithParams.top_p).toEqual(1); // top_p should be transformed correctly
     });
 
     describe('Error', () => {
@@ -175,7 +158,7 @@ describe('LobeZhipuAI', () => {
 
       it('should throw AgentRuntimeError with NoOpenAIAPIKey if no apiKey is provided', async () => {
         try {
-          await LobeZhipuAI.fromAPIKey({ apiKey: '' });
+          new LobeZhipuAI({ apiKey: '' });
         } catch (e) {
           expect(e).toEqual({ errorType: invalidErrorType });
         }
@@ -221,7 +204,7 @@ describe('LobeZhipuAI', () => {
         };
         const apiError = new OpenAI.APIError(400, errorInfo, 'module error', {});
 
-        instance = await LobeZhipuAI.fromAPIKey({
+        instance = new LobeZhipuAI({
           apiKey: 'test',
 
           baseURL: 'https://abc.com/v2',