Merge pull request #5621 from ConnectAI-E/hotfix/plugin-result

hotfix plugin result is not string #5614

README.md

@@ -12,15 +12,18 @@ One-Click to get a well-designed cross-platform ChatGPT web UI, with GPT3, GPT4
 
 一键免费部署你的跨平台私人 ChatGPT 应用, 支持 GPT3, GPT4 & Gemini Pro 模型。
 
+[![Saas][Saas-image]][saas-url]
 [![Web][Web-image]][web-url]
 [![Windows][Windows-image]][download-url]
 [![MacOS][MacOS-image]][download-url]
 [![Linux][Linux-image]][download-url]
 
-[Web App](https://app.nextchat.dev/) / [Desktop App](https://github.com/Yidadaa/ChatGPT-Next-Web/releases) / [Discord](https://discord.gg/YCkeafCafC) / [Enterprise Edition](#enterprise-edition) / [Twitter](https://twitter.com/NextChatDev)
+[NextChatAI](https://nextchat.dev/chat?utm_source=readme) / [Web App](https://app.nextchat.dev) / [Desktop App](https://github.com/Yidadaa/ChatGPT-Next-Web/releases) / [Discord](https://discord.gg/YCkeafCafC) / [Enterprise Edition](#enterprise-edition) / [Twitter](https://twitter.com/NextChatDev)
 
-[网页版](https://app.nextchat.dev/) / [客户端](https://github.com/Yidadaa/ChatGPT-Next-Web/releases) / [企业版](#%E4%BC%81%E4%B8%9A%E7%89%88) / [反馈](https://github.com/Yidadaa/ChatGPT-Next-Web/issues)
+[NextChatAI](https://nextchat.dev/chat) / [网页版](https://app.nextchat.dev) / [客户端](https://github.com/Yidadaa/ChatGPT-Next-Web/releases) / [企业版](#%E4%BC%81%E4%B8%9A%E7%89%88) / [反馈](https://github.com/Yidadaa/ChatGPT-Next-Web/issues)
 
+[saas-url]: https://nextchat.dev/chat?utm_source=readme
+[saas-image]: https://img.shields.io/badge/NextChat-Saas-green?logo=microsoftedge
 [web-url]: https://app.nextchat.dev/
 [download-url]: https://github.com/Yidadaa/ChatGPT-Next-Web/releases
 [Web-image]: https://img.shields.io/badge/Web-PWA-orange?logo=microsoftedge
@@ -60,7 +63,7 @@ For enterprise inquiries, please contact: **business@nextchat.dev**
 
 企业版咨询: **business@nextchat.dev**
 
-<img width="300" src="https://github.com/user-attachments/assets/3daeb7b6-ab63-4542-9141-2e4a12c80601">
+<img width="300" src="https://github.com/user-attachments/assets/3d4305ac-6e95-489e-884b-51d51db5f692">
 
 ## Features
 
@@ -97,6 +100,7 @@ For enterprise inquiries, please contact: **business@nextchat.dev**
 
 ## What's New
 
+- 🚀 v2.15.4 The Application supports using Tauri fetch LLM API, MORE SECURITY! [#5379](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5379)
 - 🚀 v2.15.0 Now supports Plugins! Read this: [NextChat-Awesome-Plugins](https://github.com/ChatGPTNextWeb/NextChat-Awesome-Plugins)
 - 🚀 v2.14.0 Now supports  Artifacts & SD 
 - 🚀 v2.10.1 support Google Gemini Pro model.
@@ -134,6 +138,7 @@ For enterprise inquiries, please contact: **business@nextchat.dev**
 
 ## 最新动态
 
+- 🚀 v2.15.4 客户端支持Tauri本地直接调用大模型API，更安全！[#5379](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5379)
 - 🚀 v2.15.0 现在支持插件功能了！了解更多：[NextChat-Awesome-Plugins](https://github.com/ChatGPTNextWeb/NextChat-Awesome-Plugins)
 - 🚀 v2.14.0 现在支持 Artifacts & SD 了。
 - 🚀 v2.10.1 现在支持 Gemini Pro 模型。
@@ -172,7 +177,7 @@ We recommend that you follow the steps below to re-deploy:
 
 ### Enable Automatic Updates
 
-> If you encounter a failure of Upstream Sync execution, please manually sync fork once.
+> If you encounter a failure of Upstream Sync execution, please [manually update code](./README.md#manually-updating-code).
 
 After forking the project, due to the limitations imposed by GitHub, you need to manually enable Workflows and Upstream Sync Action on the Actions page of the forked project. Once enabled, automatic updates will be scheduled every hour:
 
@@ -329,9 +334,9 @@ To control custom models, use `+` to add a custom model, use `-` to hide a model
 
 User `-all` to disable all default models, `+all` to enable all default models.
 
-For Azure: use `modelName@azure=deploymentName` to customize model name and deployment name.
-> Example: `+gpt-3.5-turbo@azure=gpt35` will show option `gpt35(Azure)` in model list.
-> If you only can use Azure model, `-all,+gpt-3.5-turbo@azure=gpt35` will `gpt35(Azure)` the only option in model list.
+For Azure: use `modelName@Azure=deploymentName` to customize model name and deployment name.
+> Example: `+gpt-3.5-turbo@Azure=gpt35` will show option `gpt35(Azure)` in model list.
+> If you only can use Azure model, `-all,+gpt-3.5-turbo@Azure=gpt35` will `gpt35(Azure)` the only option in model list.
 
 For ByteDance: use `modelName@bytedance=deploymentName` to customize model name and deployment name.
 > Example: `+Doubao-lite-4k@bytedance=ep-xxxxx-xxx` will show option `Doubao-lite-4k(ByteDance)` in model list.

README_CN.md

@@ -8,7 +8,7 @@
 
 一键免费部署你的私人 ChatGPT 网页应用，支持 GPT3, GPT4 & Gemini Pro 模型。
 
-[企业版](#%E4%BC%81%E4%B8%9A%E7%89%88) /[演示 Demo](https://chat-gpt-next-web.vercel.app/) / [反馈 Issues](https://github.com/Yidadaa/ChatGPT-Next-Web/issues) / [加入 Discord](https://discord.gg/zrhvHCr79N)
+[NextChatAI](https://nextchat.dev/chat?utm_source=readme) / [企业版](#%E4%BC%81%E4%B8%9A%E7%89%88) / [演示 Demo](https://chat-gpt-next-web.vercel.app/) / [反馈 Issues](https://github.com/Yidadaa/ChatGPT-Next-Web/issues) / [加入 Discord](https://discord.gg/zrhvHCr79N)
 
 [<img src="https://vercel.com/button" alt="Deploy on Zeabur" height="30">](https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2FChatGPTNextWeb%2FChatGPT-Next-Web&env=OPENAI_API_KEY&env=CODE&project-name=nextchat&repository-name=NextChat) [<img src="https://zeabur.com/button.svg" alt="Deploy on Zeabur" height="30">](https://zeabur.com/templates/ZBUEFA)  [<img src="https://gitpod.io/button/open-in-gitpod.svg" alt="Open in Gitpod" height="30">](https://gitpod.io/#https://github.com/Yidadaa/ChatGPT-Next-Web)
 
@@ -54,7 +54,7 @@
 
 ### 打开自动更新
 
-> 如果你遇到了 Upstream Sync 执行错误，请手动 Sync Fork 一次！
+> 如果你遇到了 Upstream Sync 执行错误，请[手动 Sync Fork 一次](./README_CN.md#手动更新代码)！
 
 当你 fork 项目之后，由于 Github 的限制，需要手动去你 fork 后的项目的 Actions 页面启用 Workflows，并启用 Upstream Sync Action，启用之后即可开启每小时定时自动更新：
 
@@ -216,9 +216,9 @@ ByteDance Api Url.
 
 用来控制模型列表，使用 `+` 增加一个模型，使用 `-` 来隐藏一个模型，使用 `模型名=展示名` 来自定义模型的展示名，用英文逗号隔开。
 
-在Azure的模式下，支持使用`modelName@azure=deploymentName`的方式配置模型名称和部署名称(deploy-name)
-> 示例：`+gpt-3.5-turbo@azure=gpt35`这个配置会在模型列表显示一个`gpt35(Azure)`的选项。
-> 如果你只能使用Azure模式，那么设置 `-all,+gpt-3.5-turbo@azure=gpt35` 则可以让对话的默认使用 `gpt35(Azure)`
+在Azure的模式下，支持使用`modelName@Azure=deploymentName`的方式配置模型名称和部署名称(deploy-name)
+> 示例：`+gpt-3.5-turbo@Azure=gpt35`这个配置会在模型列表显示一个`gpt35(Azure)`的选项。
+> 如果你只能使用Azure模式，那么设置 `-all,+gpt-3.5-turbo@Azure=gpt35` 则可以让对话的默认使用 `gpt35(Azure)`
 
 在ByteDance的模式下，支持使用`modelName@bytedance=deploymentName`的方式配置模型名称和部署名称(deploy-name)
 > 示例: `+Doubao-lite-4k@bytedance=ep-xxxxx-xxx`这个配置会在模型列表显示一个`Doubao-lite-4k(ByteDance)`的选项

README_JA.md

@@ -5,7 +5,7 @@
 
 ワンクリックで無料であなた専用の ChatGPT ウェブアプリをデプロイ。GPT3、GPT4 & Gemini Pro モデルをサポート。
 
-[企業版](#企業版) / [デモ](https://chat-gpt-next-web.vercel.app/) / [フィードバック](https://github.com/Yidadaa/ChatGPT-Next-Web/issues) / [Discordに参加](https://discord.gg/zrhvHCr79N)
+[NextChatAI](https://nextchat.dev/chat?utm_source=readme) / [企業版](#企業版) / [デモ](https://chat-gpt-next-web.vercel.app/) / [フィードバック](https://github.com/Yidadaa/ChatGPT-Next-Web/issues) / [Discordに参加](https://discord.gg/zrhvHCr79N)
 
 [<img src="https://vercel.com/button" alt="Zeaburでデプロイ" height="30">](https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2FChatGPTNextWeb%2FChatGPT-Next-Web&env=OPENAI_API_KEY&env=CODE&project-name=nextchat&repository-name=NextChat) [<img src="https://zeabur.com/button.svg" alt="Zeaburでデプロイ" height="30">](https://zeabur.com/templates/ZBUEFA)  [<img src="https://gitpod.io/button/open-in-gitpod.svg" alt="Gitpodで開く" height="30">](https://gitpod.io/#https://github.com/Yidadaa/ChatGPT-Next-Web)
 
@@ -54,7 +54,7 @@
 
 ### 自動更新を開く
 
-> Upstream Sync の実行エラーが発生した場合は、手動で Sync Fork してください！
+> Upstream Sync の実行エラーが発生した場合は、[手動で Sync Fork](./README_JA.md#手動でコードを更新する) してください！
 
 プロジェクトを fork した後、GitHub の制限により、fork 後のプロジェクトの Actions ページで Workflows を手動で有効にし、Upstream Sync Action を有効にする必要があります。有効化後、毎時の定期自動更新が可能になります：
 
@@ -207,8 +207,8 @@ ByteDance API の URL。
 
 モデルリストを管理します。`+` でモデルを追加し、`-` でモデルを非表示にし、`モデル名=表示名` でモデルの表示名をカスタマイズし、カンマで区切ります。
 
-Azure モードでは、`modelName@azure=deploymentName` 形式でモデル名とデプロイ名（deploy-name）を設定できます。
-> 例：`+gpt-3.5-turbo@azure=gpt35` この設定でモデルリストに `gpt35(Azure)` のオプションが表示されます。
+Azure モードでは、`modelName@Azure=deploymentName` 形式でモデル名とデプロイ名（deploy-name）を設定できます。
+> 例：`+gpt-3.5-turbo@Azure=gpt35` この設定でモデルリストに `gpt35(Azure)` のオプションが表示されます。
 
 ByteDance モードでは、`modelName@bytedance=deploymentName` 形式でモデル名とデプロイ名（deploy-name）を設定できます。
 > 例: `+Doubao-lite-4k@bytedance=ep-xxxxx-xxx` この設定でモデルリストに `Doubao-lite-4k(ByteDance)` のオプションが表示されます。

app/api/google.ts

@@ -23,7 +23,8 @@ export async function handle(
     });
   }
 
-  const bearToken = req.headers.get("Authorization") ?? "";
+  const bearToken =
+    req.headers.get("x-goog-api-key") || req.headers.get("Authorization") || "";
   const token = bearToken.trim().replaceAll("Bearer ", "").trim();
 
   const apiKey = token ? token : serverConfig.googleApiKey;
@@ -91,8 +92,8 @@ async function request(req: NextRequest, apiKey: string) {
     },
     10 * 60 * 1000,
   );
-  const fetchUrl = `${baseUrl}${path}?key=${apiKey}${
-    req?.nextUrl?.searchParams?.get("alt") === "sse" ? "&alt=sse" : ""
+  const fetchUrl = `${baseUrl}${path}${
+    req?.nextUrl?.searchParams?.get("alt") === "sse" ? "?alt=sse" : ""
   }`;
 
   console.log("[Fetch Url] ", fetchUrl);
@@ -100,6 +101,9 @@ async function request(req: NextRequest, apiKey: string) {
     headers: {
       "Content-Type": "application/json",
       "Cache-Control": "no-store",
+      "x-goog-api-key":
+        req.headers.get("x-goog-api-key") ||
+        (req.headers.get("Authorization") ?? "").replace("Bearer ", ""),
     },
     method: req.method,
     body: req.body,

app/api/openai.ts

@@ -6,7 +6,7 @@ import { NextRequest, NextResponse } from "next/server";
 import { auth } from "./auth";
 import { requestOpenai } from "./common";
 
-const ALLOWD_PATH = new Set(Object.values(OpenaiPath));
+const ALLOWED_PATH = new Set(Object.values(OpenaiPath));
 
 function getModels(remoteModelRes: OpenAIListModelResponse) {
   const config = getServerSideConfig();
@@ -34,7 +34,7 @@ export async function handle(
 
   const subpath = params.path.join("/");
 
-  if (!ALLOWD_PATH.has(subpath)) {
+  if (!ALLOWED_PATH.has(subpath)) {
     console.log("[OpenAI Route] forbidden path ", subpath);
     return NextResponse.json(
       {

app/client/api.ts

@@ -231,7 +231,7 @@ export function getHeaders(ignoreHeaders: boolean = false) {
 
   function getConfig() {
     const modelConfig = chatStore.currentSession().mask.modelConfig;
-    const isGoogle = modelConfig.providerName == ServiceProvider.Google;
+    const isGoogle = modelConfig.providerName === ServiceProvider.Google;
     const isAzure = modelConfig.providerName === ServiceProvider.Azure;
     const isAnthropic = modelConfig.providerName === ServiceProvider.Anthropic;
     const isBaidu = modelConfig.providerName == ServiceProvider.Baidu;
@@ -272,7 +272,13 @@ export function getHeaders(ignoreHeaders: boolean = false) {
   }
 
   function getAuthHeader(): string {
-    return isAzure ? "api-key" : isAnthropic ? "x-api-key" : "Authorization";
+    return isAzure
+      ? "api-key"
+      : isAnthropic
+      ? "x-api-key"
+      : isGoogle
+      ? "x-goog-api-key"
+      : "Authorization";
   }
 
   const {
@@ -283,14 +289,15 @@ export function getHeaders(ignoreHeaders: boolean = false) {
     apiKey,
     isEnabledAccessControl,
   } = getConfig();
-  // when using google api in app, not set auth header
-  if (isGoogle && clientConfig?.isApp) return headers;
   // when using baidu api in app, not set auth header
   if (isBaidu && clientConfig?.isApp) return headers;
 
   const authHeader = getAuthHeader();
 
-  const bearerToken = getBearerToken(apiKey, isAzure || isAnthropic);
+  const bearerToken = getBearerToken(
+    apiKey,
+    isAzure || isAnthropic || isGoogle,
+  );
 
   if (bearerToken) {
     headers[authHeader] = bearerToken;

app/client/platforms/alibaba.ts

@@ -23,6 +23,7 @@ import {
 import { prettyObject } from "@/app/utils/format";
 import { getClientConfig } from "@/app/config/client";
 import { getMessageTextContent } from "@/app/utils";
+import { fetch } from "@/app/utils/stream";
 
 export interface OpenAIListModelResponse {
   object: string;
@@ -178,6 +179,7 @@ export class QwenApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/client/platforms/anthropic.ts

@@ -8,7 +8,7 @@ import {
   ChatMessageTool,
 } from "@/app/store";
 import { getClientConfig } from "@/app/config/client";
-import { DEFAULT_API_HOST } from "@/app/constant";
+import { ANTHROPIC_BASE_URL } from "@/app/constant";
 import { getMessageTextContent, isVisionModel } from "@/app/utils";
 import { preProcessImageContent, stream } from "@/app/utils/chat";
 import { cloudflareAIGatewayUrl } from "@/app/utils/cloudflare";
@@ -388,9 +388,7 @@ export class ClaudeApi implements LLMApi {
     if (baseUrl.trim().length === 0) {
       const isApp = !!getClientConfig()?.isApp;
 
-      baseUrl = isApp
-        ? DEFAULT_API_HOST + "/api/proxy/anthropic"
-        : ApiPath.Anthropic;
+      baseUrl = isApp ? ANTHROPIC_BASE_URL : ApiPath.Anthropic;
     }
 
     if (!baseUrl.startsWith("http") && !baseUrl.startsWith("/api")) {

app/client/platforms/baidu.ts

@@ -24,6 +24,7 @@ import {
 import { prettyObject } from "@/app/utils/format";
 import { getClientConfig } from "@/app/config/client";
 import { getMessageTextContent } from "@/app/utils";
+import { fetch } from "@/app/utils/stream";
 
 export interface OpenAIListModelResponse {
   object: string;
@@ -197,6 +198,7 @@ export class ErnieApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/client/platforms/bytedance.ts

@@ -23,6 +23,7 @@ import {
 import { prettyObject } from "@/app/utils/format";
 import { getClientConfig } from "@/app/config/client";
 import { getMessageTextContent } from "@/app/utils";
+import { fetch } from "@/app/utils/stream";
 
 export interface OpenAIListModelResponse {
   object: string;
@@ -165,6 +166,7 @@ export class DoubaoApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/client/platforms/google.ts

@@ -9,7 +9,7 @@ import {
 } from "../api";
 import { useAccessStore, useAppConfig, useChatStore } from "@/app/store";
 import { getClientConfig } from "@/app/config/client";
-import { DEFAULT_API_HOST } from "@/app/constant";
+import { GEMINI_BASE_URL } from "@/app/constant";
 import Locale from "../../locales";
 import {
   EventStreamContentType,
@@ -22,6 +22,7 @@ import {
   isVisionModel,
 } from "@/app/utils";
 import { preProcessImageContent } from "@/app/utils/chat";
+import { fetch } from "@/app/utils/stream";
 
 export class GeminiProApi implements LLMApi {
   path(path: string): string {
@@ -34,7 +35,7 @@ export class GeminiProApi implements LLMApi {
 
     const isApp = !!getClientConfig()?.isApp;
     if (baseUrl.length === 0) {
-      baseUrl = isApp ? DEFAULT_API_HOST + `/api/proxy/google` : ApiPath.Google;
+      baseUrl = isApp ? GEMINI_BASE_URL : ApiPath.Google;
     }
     if (baseUrl.endsWith("/")) {
       baseUrl = baseUrl.slice(0, baseUrl.length - 1);
@@ -48,10 +49,6 @@ export class GeminiProApi implements LLMApi {
     let chatPath = [baseUrl, path].join("/");
 
     chatPath += chatPath.includes("?") ? "&alt=sse" : "?alt=sse";
-    // if chatPath.startsWith('http') then add key in query string
-    if (chatPath.startsWith("http") && accessStore.googleApiKey) {
-      chatPath += `&key=${accessStore.googleApiKey}`;
-    }
     return chatPath;
   }
   extractMessage(res: any) {
@@ -217,6 +214,7 @@ export class GeminiProApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/client/platforms/iflytek.ts

@@ -1,7 +1,7 @@
 "use client";
 import {
   ApiPath,
-  DEFAULT_API_HOST,
+  IFLYTEK_BASE_URL,
   Iflytek,
   REQUEST_TIMEOUT_MS,
 } from "@/app/constant";
@@ -22,6 +22,7 @@ import {
 import { prettyObject } from "@/app/utils/format";
 import { getClientConfig } from "@/app/config/client";
 import { getMessageTextContent } from "@/app/utils";
+import { fetch } from "@/app/utils/stream";
 
 import { RequestPayload } from "./openai";
 
@@ -40,7 +41,7 @@ export class SparkApi implements LLMApi {
     if (baseUrl.length === 0) {
       const isApp = !!getClientConfig()?.isApp;
       const apiPath = ApiPath.Iflytek;
-      baseUrl = isApp ? DEFAULT_API_HOST + "/proxy" + apiPath : apiPath;
+      baseUrl = isApp ? IFLYTEK_BASE_URL : apiPath;
     }
 
     if (baseUrl.endsWith("/")) {
@@ -149,6 +150,7 @@ export class SparkApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/client/platforms/moonshot.ts

@@ -2,7 +2,7 @@
 // azure and openai, using same models. so using same LLMApi.
 import {
   ApiPath,
-  DEFAULT_API_HOST,
+  MOONSHOT_BASE_URL,
   Moonshot,
   REQUEST_TIMEOUT_MS,
 } from "@/app/constant";
@@ -40,7 +40,7 @@ export class MoonshotApi implements LLMApi {
     if (baseUrl.length === 0) {
       const isApp = !!getClientConfig()?.isApp;
       const apiPath = ApiPath.Moonshot;
-      baseUrl = isApp ? DEFAULT_API_HOST + "/proxy" + apiPath : apiPath;
+      baseUrl = isApp ? MOONSHOT_BASE_URL : apiPath;
     }
 
     if (baseUrl.endsWith("/")) {

app/client/platforms/openai.ts

@@ -2,7 +2,7 @@
 // azure and openai, using same models. so using same LLMApi.
 import {
   ApiPath,
-  DEFAULT_API_HOST,
+  OPENAI_BASE_URL,
   DEFAULT_MODELS,
   OpenaiPath,
   Azure,
@@ -98,7 +98,7 @@ export class ChatGPTApi implements LLMApi {
     if (baseUrl.length === 0) {
       const isApp = !!getClientConfig()?.isApp;
       const apiPath = isAzure ? ApiPath.Azure : ApiPath.OpenAI;
-      baseUrl = isApp ? DEFAULT_API_HOST + "/proxy" + apiPath : apiPath;
+      baseUrl = isApp ? OPENAI_BASE_URL : apiPath;
     }
 
     if (baseUrl.endsWith("/")) {

app/client/platforms/tencent.ts

@@ -1,5 +1,5 @@
 "use client";
-import { ApiPath, DEFAULT_API_HOST, REQUEST_TIMEOUT_MS } from "@/app/constant";
+import { ApiPath, TENCENT_BASE_URL, REQUEST_TIMEOUT_MS } from "@/app/constant";
 import { useAccessStore, useAppConfig, useChatStore } from "@/app/store";
 
 import {
@@ -22,6 +22,7 @@ import mapKeys from "lodash-es/mapKeys";
 import mapValues from "lodash-es/mapValues";
 import isArray from "lodash-es/isArray";
 import isObject from "lodash-es/isObject";
+import { fetch } from "@/app/utils/stream";
 
 export interface OpenAIListModelResponse {
   object: string;
@@ -70,9 +71,7 @@ export class HunyuanApi implements LLMApi {
 
     if (baseUrl.length === 0) {
       const isApp = !!getClientConfig()?.isApp;
-      baseUrl = isApp
-        ? DEFAULT_API_HOST + "/api/proxy/tencent"
-        : ApiPath.Tencent;
+      baseUrl = isApp ? TENCENT_BASE_URL : ApiPath.Tencent;
     }
 
     if (baseUrl.endsWith("/")) {
@@ -179,6 +178,7 @@ export class HunyuanApi implements LLMApi {
         controller.signal.onabort = finish;
 
         fetchEventSource(chatPath, {
+          fetch: fetch as any,
           ...chatPayload,
           async onopen(res) {
             clearTimeout(requestTimeoutId);

app/components/chat.tsx

@@ -1815,6 +1815,7 @@ function _Chat() {
                       {message?.tools?.map((tool) => (
                         <div
                           key={tool.id}
+                          title={tool?.errorMsg}
                           className={styles["chat-message-tool"]}
                         >
                           {tool.isError === false ? (

app/components/markdown.tsx

@@ -207,23 +207,6 @@ function CustomCode(props: { children: any; className?: string }) {
   );
 }
 
-function escapeDollarNumber(text: string) {
-  let escapedText = "";
-
-  for (let i = 0; i < text.length; i += 1) {
-    let char = text[i];
-    const nextChar = text[i + 1] || " ";
-
-    if (char === "$" && nextChar >= "0" && nextChar <= "9") {
-      char = "\\$";
-    }
-
-    escapedText += char;
-  }
-
-  return escapedText;
-}
-
 function escapeBrackets(text: string) {
   const pattern =
     /(```[\s\S]*?```|`.*?`)|\\\[([\s\S]*?[^\\])\\\]|\\\((.*?)\\\)/g;
@@ -252,7 +235,7 @@ function tryWrapHtmlCode(text: string) {
       },
     )
     .replace(
-      /(<\/body>)([\r\n\s]*?)(<\/html>)([\n\r]*?)([`]*?)([\n\r]*?)/g,
+      /(<\/body>)([\r\n\s]*?)(<\/html>)([\n\r]*)([`]*)([\n\r]*?)/g,
       (match, bodyEnd, space, htmlEnd, newLine, quoteEnd) => {
         return !quoteEnd ? bodyEnd + space + htmlEnd + "\n```\n" : match;
       },
@@ -261,7 +244,7 @@ function tryWrapHtmlCode(text: string) {
 
 function _MarkDownContent(props: { content: string }) {
   const escapedContent = useMemo(() => {
-    return tryWrapHtmlCode(escapeBrackets(escapeDollarNumber(props.content)));
+    return tryWrapHtmlCode(escapeBrackets(props.content));
   }, [props.content]);
 
   return (

app/constant.ts

@@ -11,7 +11,6 @@ export const RUNTIME_CONFIG_DOM = "danger-runtime-config";
 
 export const STABILITY_BASE_URL = "https://api.stability.ai";
 
-export const DEFAULT_API_HOST = "https://api.nextchat.dev";
 export const OPENAI_BASE_URL = "https://api.openai.com";
 export const ANTHROPIC_BASE_URL = "https://api.anthropic.com";
 

app/locales/tw.ts

@@ -8,12 +8,12 @@ const tw = {
   Error: {
     Unauthorized: isApp
       ? `😆 對話遇到了一些問題，不用慌:
-    \\ 1️⃣ 想要零配置開箱即用，[點擊這裡立刻開啟對話 🚀](${SAAS_CHAT_UTM_URL})
-    \\ 2️⃣ 如果你想消耗自己的 OpenAI 資源，點擊[這裡](/#/settings)修改設定 ⚙️`
+    \\ 1️⃣ 想要無須設定開箱即用，[點選這裡立刻開啟對話 🚀](${SAAS_CHAT_UTM_URL})
+    \\ 2️⃣ 如果你想消耗自己的 OpenAI 資源，點選[這裡](/#/settings)修改設定 ⚙️`
       : `😆 對話遇到了一些問題，不用慌:
-    \ 1️⃣ 想要零配置開箱即用，[點擊這裡立刻開啟對話 🚀](${SAAS_CHAT_UTM_URL})
-    \ 2️⃣ 如果你正在使用私有部署版本，點擊[這裡](/#/auth)輸入訪問秘鑰 🔑
-    \ 3️⃣ 如果你想消耗自己的 OpenAI 資源，點擊[這裡](/#/settings)修改設定 ⚙️
+    \ 1️⃣ 想要無須設定開箱即用，[點選這裡立刻開啟對話 🚀](${SAAS_CHAT_UTM_URL})
+    \ 2️⃣ 如果你正在使用私有部署版本，點選[這裡](/#/auth)輸入存取金鑰 🔑
+    \ 3️⃣ 如果你想消耗自己的 OpenAI 資源，點選[這裡](/#/settings)修改設定 ⚙️
  `,
   },
 
@@ -25,9 +25,9 @@ const tw = {
     Confirm: "確認",
     Later: "稍候再說",
     Return: "返回",
-    SaasTips: "配置太麻煩，想要立即使用",
+    SaasTips: "設定太麻煩，想要立即使用",
     TopTips:
-      "🥳 NextChat AI 首發優惠，立刻解鎖 OpenAI o1, GPT-4o, Claude-3.5 等最新大模型",
+      "🥳 NextChat AI 首發優惠，立刻解鎖 OpenAI o1, GPT-4o, Claude-3.5 等最新的大型語言模型",
   },
   ChatItem: {
     ChatItemCount: (count: number) => `${count} 則對話`,
@@ -53,8 +53,8 @@ const tw = {
       PinToastAction: "檢視",
       Delete: "刪除",
       Edit: "編輯",
-      RefreshTitle: "刷新標題",
-      RefreshToast: "已發送刷新標題請求",
+      RefreshTitle: "重新整理標題",
+      RefreshToast: "已傳送重新整理標題請求",
     },
     Commands: {
       new: "新建聊天",
@@ -95,10 +95,10 @@ const tw = {
     IsContext: "預設提示詞",
     ShortcutKey: {
       Title: "鍵盤快捷方式",
-      newChat: "打開新聊天",
+      newChat: "開啟新聊天",
       focusInput: "聚焦輸入框",
       copyLastMessage: "複製最後一個回覆",
-      copyLastCode: "複製最後一個代碼塊",
+      copyLastCode: "複製最後一個程式碼區塊",
       showShortcutKey: "顯示快捷方式",
     },
   },
@@ -174,9 +174,9 @@ const tw = {
       SubTitle: "聊天內容的字型大小",
     },
     FontFamily: {
-      Title: "聊天字體",
-      SubTitle: "聊天內容的字體，若置空則應用全局默認字體",
-      Placeholder: "字體名稱",
+      Title: "聊天字型",
+      SubTitle: "聊天內容的字型，若留空則套用全域預設字型",
+      Placeholder: "字型名稱",
     },
     InjectSystemPrompts: {
       Title: "匯入系統提示",
@@ -301,8 +301,8 @@ const tw = {
         Title: "使用 NextChat AI",
         Label: "(性價比最高的方案)",
         SubTitle:
-          "由 NextChat 官方維護，零配置開箱即用，支持 OpenAI o1、GPT-4o、Claude-3.5 等最新大模型",
-        ChatNow: "立刻對話",
+          "由 NextChat 官方維護，無須設定開箱即用，支援 OpenAI o1、GPT-4o、Claude-3.5 等最新的大型語言模型",
+        ChatNow: "立刻開始對話",
       },
 
       AccessCode: {
@@ -485,18 +485,18 @@ const tw = {
     },
   },
   SearchChat: {
-    Name: "搜索",
+    Name: "搜尋",
     Page: {
-      Title: "搜索聊天記錄",
-      Search: "輸入搜索關鍵詞",
+      Title: "搜尋聊天記錄",
+      Search: "輸入搜尋關鍵詞",
       NoResult: "沒有找到結果",
-      NoData: "沒有數據",
-      Loading: "加載中",
+      NoData: "沒有資料",
+      Loading: "載入中",
 
       SubTitle: (count: number) => `找到 ${count} 條結果`,
     },
     Item: {
-      View: "查看",
+      View: "檢視",
     },
   },
   NewChat: {

app/store/access.ts

@@ -1,9 +1,18 @@
 import {
-  ApiPath,
-  DEFAULT_API_HOST,
   GoogleSafetySettingsThreshold,
   ServiceProvider,
   StoreKey,
+  ApiPath,
+  OPENAI_BASE_URL,
+  ANTHROPIC_BASE_URL,
+  GEMINI_BASE_URL,
+  BAIDU_BASE_URL,
+  BYTEDANCE_BASE_URL,
+  ALIBABA_BASE_URL,
+  TENCENT_BASE_URL,
+  MOONSHOT_BASE_URL,
+  STABILITY_BASE_URL,
+  IFLYTEK_BASE_URL,
 } from "../constant";
 import { getHeaders } from "../client/api";
 import { getClientConfig } from "../config/client";
@@ -15,45 +24,25 @@ let fetchState = 0; // 0 not fetch, 1 fetching, 2 done
 
 const isApp = getClientConfig()?.buildMode === "export";
 
-const DEFAULT_OPENAI_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/openai"
-  : ApiPath.OpenAI;
+const DEFAULT_OPENAI_URL = isApp ? OPENAI_BASE_URL : ApiPath.OpenAI;
 
-const DEFAULT_GOOGLE_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/google"
-  : ApiPath.Google;
+const DEFAULT_GOOGLE_URL = isApp ? GEMINI_BASE_URL : ApiPath.Google;
 
-const DEFAULT_ANTHROPIC_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/anthropic"
-  : ApiPath.Anthropic;
+const DEFAULT_ANTHROPIC_URL = isApp ? ANTHROPIC_BASE_URL : ApiPath.Anthropic;
 
-const DEFAULT_BAIDU_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/baidu"
-  : ApiPath.Baidu;
+const DEFAULT_BAIDU_URL = isApp ? BAIDU_BASE_URL : ApiPath.Baidu;
 
-const DEFAULT_BYTEDANCE_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/bytedance"
-  : ApiPath.ByteDance;
+const DEFAULT_BYTEDANCE_URL = isApp ? BYTEDANCE_BASE_URL : ApiPath.ByteDance;
 
-const DEFAULT_ALIBABA_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/alibaba"
-  : ApiPath.Alibaba;
+const DEFAULT_ALIBABA_URL = isApp ? ALIBABA_BASE_URL : ApiPath.Alibaba;
 
-const DEFAULT_TENCENT_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/tencent"
-  : ApiPath.Tencent;
+const DEFAULT_TENCENT_URL = isApp ? TENCENT_BASE_URL : ApiPath.Tencent;
 
-const DEFAULT_MOONSHOT_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/moonshot"
-  : ApiPath.Moonshot;
+const DEFAULT_MOONSHOT_URL = isApp ? MOONSHOT_BASE_URL : ApiPath.Moonshot;
 
-const DEFAULT_STABILITY_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/stability"
-  : ApiPath.Stability;
+const DEFAULT_STABILITY_URL = isApp ? STABILITY_BASE_URL : ApiPath.Stability;
 
-const DEFAULT_IFLYTEK_URL = isApp
-  ? DEFAULT_API_HOST + "/api/proxy/iflytek"
-  : ApiPath.Iflytek;
+const DEFAULT_IFLYTEK_URL = isApp ? IFLYTEK_BASE_URL : ApiPath.Iflytek;
 
 const DEFAULT_ACCESS_STATE = {
   accessCode: "",
@@ -211,10 +200,13 @@ export const useAccessStore = createPersistStore(
       })
         .then((res) => res.json())
         .then((res) => {
-          // Set default model from env request
-          let defaultModel = res.defaultModel ?? "";
-          if (defaultModel !== "")
-            DEFAULT_CONFIG.modelConfig.model = defaultModel;
+          const defaultModel = res.defaultModel ?? "";
+          if (defaultModel !== "") {
+            const [model, providerName] = defaultModel.split("@");
+            DEFAULT_CONFIG.modelConfig.model = model;
+            DEFAULT_CONFIG.modelConfig.providerName = providerName;
+          }
+
           return res;
         })
         .then((res: DangerConfig) => {

app/store/chat.ts

@@ -16,6 +16,9 @@ import {
   DEFAULT_SYSTEM_TEMPLATE,
   KnowledgeCutOffDate,
   StoreKey,
+  SUMMARIZE_MODEL,
+  GEMINI_SUMMARIZE_MODEL,
+  ServiceProvider,
 } from "../constant";
 import Locale, { getLang } from "../locales";
 import { isDalle3, safeLocalStorage } from "../utils";
@@ -23,6 +26,8 @@ import { prettyObject } from "../utils/format";
 import { createPersistStore } from "../utils/store";
 import { estimateTokenLength } from "../utils/token";
 import { ModelConfig, ModelType, useAppConfig } from "./config";
+import { useAccessStore } from "./access";
+import { collectModelsWithDefaultModel } from "../utils/model";
 import { createEmptyMask, Mask } from "./mask";
 
 const localStorage = safeLocalStorage();
@@ -37,6 +42,7 @@ export type ChatMessageTool = {
   };
   content?: string;
   isError?: boolean;
+  errorMsg?: string;
 };
 
 export type ChatMessage = RequestMessage & {
@@ -102,6 +108,35 @@ function createEmptySession(): ChatSession {
   };
 }
 
+function getSummarizeModel(
+  currentModel: string,
+  providerName: string,
+): string[] {
+  // if it is using gpt-* models, force to use 4o-mini to summarize
+  if (currentModel.startsWith("gpt") || currentModel.startsWith("chatgpt")) {
+    const configStore = useAppConfig.getState();
+    const accessStore = useAccessStore.getState();
+    const allModel = collectModelsWithDefaultModel(
+      configStore.models,
+      [configStore.customModels, accessStore.customModels].join(","),
+      accessStore.defaultModel,
+    );
+    const summarizeModel = allModel.find(
+      (m) => m.name === SUMMARIZE_MODEL && m.available,
+    );
+    if (summarizeModel) {
+      return [
+        summarizeModel.name,
+        summarizeModel.provider?.providerName as string,
+      ];
+    }
+  }
+  if (currentModel.startsWith("gemini")) {
+    return [GEMINI_SUMMARIZE_MODEL, ServiceProvider.Google];
+  }
+  return [currentModel, providerName];
+}
+
 function countMessages(msgs: ChatMessage[]) {
   return msgs.reduce(
     (pre, cur) => pre + estimateTokenLength(getMessageTextContent(cur)),
@@ -578,8 +613,14 @@ export const useChatStore = createPersistStore(
           return;
         }
 
-        const providerName = modelConfig.compressProviderName;
-        const api: ClientApi = getClientApi(providerName);
+        // if not config compressModel, then using getSummarizeModel
+        const [model, providerName] = modelConfig.compressModel
+          ? [modelConfig.compressModel, modelConfig.compressProviderName]
+          : getSummarizeModel(
+              session.mask.modelConfig.model,
+              session.mask.modelConfig.providerName,
+            );
+        const api: ClientApi = getClientApi(providerName as ServiceProvider);
 
         // remove error messages if any
         const messages = session.messages;
@@ -610,7 +651,7 @@ export const useChatStore = createPersistStore(
           api.llm.chat({
             messages: topicMessages,
             config: {
-              model: modelConfig.compressModel,
+              model,
               stream: false,
               providerName,
             },
@@ -674,7 +715,8 @@ export const useChatStore = createPersistStore(
             config: {
               ...modelcfg,
               stream: true,
-              model: modelConfig.compressModel,
+              model,
+              providerName,
             },
             onUpdate(message) {
               session.memoryPrompt = message;
@@ -727,7 +769,7 @@ export const useChatStore = createPersistStore(
   },
   {
     name: StoreKey.Chat,
-    version: 3.2,
+    version: 3.3,
     migrate(persistedState, version) {
       const state = persistedState as any;
       const newState = JSON.parse(
@@ -783,6 +825,14 @@ export const useChatStore = createPersistStore(
             config.modelConfig.compressProviderName;
         });
       }
+      // revert default summarize model for every session
+      if (version < 3.3) {
+        newState.sessions.forEach((s) => {
+          const config = useAppConfig.getState();
+          s.mask.modelConfig.compressModel = "";
+          s.mask.modelConfig.compressProviderName = "";
+        });
+      }
 
       return newState as any;
     },

app/store/config.ts

@@ -71,8 +71,8 @@ export const DEFAULT_CONFIG = {
     sendMemory: true,
     historyMessageCount: 4,
     compressMessageLengthThreshold: 1000,
-    compressModel: "gpt-4o-mini" as ModelType,
-    compressProviderName: "OpenAI" as ServiceProvider,
+    compressModel: "",
+    compressProviderName: "",
     enableInjectSystemPrompts: true,
     template: config?.template ?? DEFAULT_INPUT_TEMPLATE,
     size: "1024x1024" as DalleSize,
@@ -178,7 +178,7 @@ export const useAppConfig = createPersistStore(
   }),
   {
     name: StoreKey.Config,
-    version: 4,
+    version: 4.1,
 
     merge(persistedState, currentState) {
       const state = persistedState as ChatConfig | undefined;
@@ -231,7 +231,7 @@ export const useAppConfig = createPersistStore(
             : config?.template ?? DEFAULT_INPUT_TEMPLATE;
       }
 
-      if (version < 4) {
+      if (version < 4.1) {
         state.modelConfig.compressModel =
           DEFAULT_CONFIG.modelConfig.compressModel;
         state.modelConfig.compressProviderName =

app/store/plugin.ts

@@ -4,10 +4,10 @@ import { nanoid } from "nanoid";
 import { createPersistStore } from "../utils/store";
 import { getClientConfig } from "../config/client";
 import yaml from "js-yaml";
-import { adapter } from "../utils";
+import { adapter, getOperationId } from "../utils";
 import { useAccessStore } from "./access";
 
-const isApp = getClientConfig()?.isApp;
+const isApp = getClientConfig()?.isApp !== false;
 
 export type Plugin = {
   id: string;
@@ -116,7 +116,7 @@ export const FunctionToolService = {
         return {
           type: "function",
           function: {
-            name: o.operationId,
+            name: getOperationId(o),
             description: o.description || o.summary,
             parameters: parameters,
           },
@@ -124,7 +124,7 @@ export const FunctionToolService = {
       }),
       funcs: operations.reduce((s, o) => {
         // @ts-ignore
-        s[o.operationId] = function (args) {
+        s[getOperationId(o)] = function (args) {
           const parameters: Record<string, any> = {};
           if (o.parameters instanceof Array) {
             o.parameters.forEach((p) => {
@@ -139,8 +139,8 @@ export const FunctionToolService = {
           } else if (authLocation == "body") {
             args[headerName] = tokenValue;
           }
-          // @ts-ignore
-          return api.client[o.operationId](
+          // @ts-ignore if o.operationId is null, then using o.path and o.method
+          return api.client.paths[o.path][o.method](
             parameters,
             args,
             api.axiosConfigDefaults,

app/store/sync.ts

@@ -12,7 +12,6 @@ import { downloadAs, readFromFile } from "../utils";
 import { showToast } from "../components/ui-lib";
 import Locale from "../locales";
 import { createSyncClient, ProviderType } from "../utils/cloud";
-import { corsPath } from "../utils/cors";
 
 export interface WebDavConfig {
   server: string;
@@ -26,7 +25,7 @@ export type SyncStore = GetStoreState<typeof useSyncStore>;
 const DEFAULT_SYNC_STATE = {
   provider: ProviderType.WebDAV,
   useProxy: true,
-  proxyUrl: corsPath(ApiPath.Cors),
+  proxyUrl: ApiPath.Cors as string,
 
   webdav: {
     endpoint: "",

app/utils.ts

@@ -2,8 +2,9 @@ import { useEffect, useState } from "react";
 import { showToast } from "./components/ui-lib";
 import Locale from "./locales";
 import { RequestMessage } from "./client/api";
-import { ServiceProvider, REQUEST_TIMEOUT_MS } from "./constant";
-import { fetch as tauriFetch, ResponseType } from "@tauri-apps/api/http";
+import { ServiceProvider } from "./constant";
+// import { fetch as tauriFetch, ResponseType } from "@tauri-apps/api/http";
+import { fetch as tauriStreamFetch } from "./utils/stream";
 
 export function trimTopic(topic: string) {
   // Fix an issue where double quotes still show in the Indonesian language
@@ -292,30 +293,23 @@ export function fetch(
   options?: Record<string, unknown>,
 ): Promise<any> {
   if (window.__TAURI__) {
-    const payload = options?.body || options?.data;
-    return tauriFetch(url, {
-      ...options,
-      body:
-        payload &&
-        ({
-          type: "Text",
-          payload,
-        } as any),
-      timeout: ((options?.timeout as number) || REQUEST_TIMEOUT_MS) / 1000,
-      responseType:
-        options?.responseType == "text" ? ResponseType.Text : ResponseType.JSON,
-    } as any);
+    return tauriStreamFetch(url, options);
   }
   return window.fetch(url, options);
 }
 
 export function adapter(config: Record<string, unknown>) {
-  const { baseURL, url, params, ...rest } = config;
+  const { baseURL, url, params, data: body, ...rest } = config;
   const path = baseURL ? `${baseURL}${url}` : url;
   const fetchUrl = params
     ? `${path}?${new URLSearchParams(params as any).toString()}`
     : path;
-  return fetch(fetchUrl as string, { ...rest, responseType: "text" });
+  return fetch(fetchUrl as string, { ...rest, body }).then((res) => {
+    const { status, headers, statusText } = res;
+    return res
+      .text()
+      .then((data: string) => ({ status, statusText, headers, data }));
+  });
 }
 
 export function safeLocalStorage(): {
@@ -377,3 +371,15 @@ export function safeLocalStorage(): {
     },
   };
 }
+
+export function getOperationId(operation: {
+  operationId?: string;
+  method: string;
+  path: string;
+}) {
+  // pattern '^[a-zA-Z0-9_-]+$'
+  return (
+    operation?.operationId ||
+    `${operation.method.toUpperCase()}${operation.path.replaceAll("/", "_")}`
+  );
+}

app/utils/chat.ts

@@ -10,6 +10,7 @@ import {
   fetchEventSource,
 } from "@fortaine/fetch-event-source";
 import { prettyObject } from "./format";
+import { fetch as tauriFetch } from "./stream";
 
 export function compressImage(file: Blob, maxSize: number): Promise<string> {
   return new Promise((resolve, reject) => {
@@ -221,7 +222,12 @@ export function stream(
               ),
             )
               .then((res) => {
-                const content = JSON.stringify(res.data);
+                let content = res.data || res?.statusText;
+                // hotfix #5614
+                content =
+                  typeof content === "string"
+                    ? content
+                    : JSON.stringify(content);
                 if (res.status >= 300) {
                   return Promise.reject(content);
                 }
@@ -236,7 +242,11 @@ export function stream(
                 return content;
               })
               .catch((e) => {
-                options?.onAfterTool?.({ ...tool, isError: true });
+                options?.onAfterTool?.({
+                  ...tool,
+                  isError: true,
+                  errorMsg: e.toString(),
+                });
                 return e.toString();
               })
               .then((content) => ({
@@ -287,6 +297,7 @@ export function stream(
       REQUEST_TIMEOUT_MS,
     );
     fetchEventSource(chatPath, {
+      fetch: tauriFetch as any,
       ...chatPayload,
       async onopen(res) {
         clearTimeout(requestTimeoutId);

app/utils/cors.ts

@@ -1,19 +0,0 @@
-import { getClientConfig } from "../config/client";
-import { DEFAULT_API_HOST } from "../constant";
-
-export function corsPath(path: string) {
-  const baseUrl = getClientConfig()?.isApp ? `${DEFAULT_API_HOST}` : "";
-
-  if (baseUrl === "" && path === "") {
-    return "";
-  }
-  if (!path.startsWith("/")) {
-    path = "/" + path;
-  }
-
-  if (!path.endsWith("/")) {
-    path += "/";
-  }
-
-  return `${baseUrl}${path}`;
-}

app/utils/stream.ts

@@ -0,0 +1,107 @@
+// using tauri command to send request
+// see src-tauri/src/stream.rs, and src-tauri/src/main.rs
+// 1. invoke('stream_fetch', {url, method, headers, body}), get response with headers.
+// 2. listen event: `stream-response` multi times to get body
+
+type ResponseEvent = {
+  id: number;
+  payload: {
+    request_id: number;
+    status?: number;
+    chunk?: number[];
+  };
+};
+
+type StreamResponse = {
+  request_id: number;
+  status: number;
+  status_text: string;
+  headers: Record<string, string>;
+};
+
+export function fetch(url: string, options?: RequestInit): Promise<any> {
+  if (window.__TAURI__) {
+    const {
+      signal,
+      method = "GET",
+      headers: _headers = {},
+      body = [],
+    } = options || {};
+    let unlisten: Function | undefined;
+    let setRequestId: Function | undefined;
+    const requestIdPromise = new Promise((resolve) => (setRequestId = resolve));
+    const ts = new TransformStream();
+    const writer = ts.writable.getWriter();
+
+    let closed = false;
+    const close = () => {
+      if (closed) return;
+      closed = true;
+      unlisten && unlisten();
+      writer.ready.then(() => {
+        writer.close().catch((e) => console.error(e));
+      });
+    };
+
+    if (signal) {
+      signal.addEventListener("abort", () => close());
+    }
+    // @ts-ignore 2. listen response multi times, and write to Response.body
+    window.__TAURI__.event
+      .listen("stream-response", (e: ResponseEvent) =>
+        requestIdPromise.then((request_id) => {
+          const { request_id: rid, chunk, status } = e?.payload || {};
+          if (request_id != rid) {
+            return;
+          }
+          if (chunk) {
+            writer.ready.then(() => {
+              writer.write(new Uint8Array(chunk));
+            });
+          } else if (status === 0) {
+            // end of body
+            close();
+          }
+        }),
+      )
+      .then((u: Function) => (unlisten = u));
+
+    const headers: Record<string, string> = {
+      Accept: "application/json, text/plain, */*",
+      "Accept-Language": "en-US,en;q=0.9,zh-CN;q=0.8,zh;q=0.7",
+      "User-Agent": navigator.userAgent,
+    };
+    for (const item of new Headers(_headers || {})) {
+      headers[item[0]] = item[1];
+    }
+    return window.__TAURI__
+      .invoke("stream_fetch", {
+        method: method.toUpperCase(),
+        url,
+        headers,
+        // TODO FormData
+        body:
+          typeof body === "string"
+            ? Array.from(new TextEncoder().encode(body))
+            : [],
+      })
+      .then((res: StreamResponse) => {
+        const { request_id, status, status_text: statusText, headers } = res;
+        setRequestId?.(request_id);
+        const response = new Response(ts.readable, {
+          status,
+          statusText,
+          headers,
+        });
+        if (status >= 300) {
+          setTimeout(close, 100);
+        }
+        return response;
+      })
+      .catch((e) => {
+        console.error("stream error", e);
+        throw e;
+      });
+  }
+  return window.fetch(url, options);
+}

src-tauri/Cargo.lock

@@ -348,9 +348,9 @@ checksum = "14c189c53d098945499cdfa7ecc63567cf3886b3332b312a5b4585d8d3a6a610"
 
 [[package]]
 name = "bytes"
-version = "1.4.0"
+version = "1.7.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "89b2fd2a0dcf38d7971e2194b6b6eebab45ae01067456a7fd93d5547a61b70be"
+checksum = "428d9aa8fbc0670b7b8d6030a7fadd0f86151cae55e4dbbece15f3780a3dfaf3"
 dependencies = [
  "serde",
 ]
@@ -942,9 +942,9 @@ checksum = "00b0228411908ca8685dba7fc2cdd70ec9990a6e753e89b6ac91a84c40fbaf4b"
 
 [[package]]
 name = "form_urlencoded"
-version = "1.1.0"
+version = "1.2.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a9c384f161156f5260c24a097c56119f9be8c798586aecc13afbcbe7b7e26bf8"
+checksum = "e13624c2627564efccf4934284bdd98cbaa14e79b0b5a141218e507b3a823456"
 dependencies = [
  "percent-encoding",
 ]
@@ -970,9 +970,9 @@ dependencies = [
 
 [[package]]
 name = "futures-core"
-version = "0.3.28"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4bca583b7e26f571124fe5b7561d49cb2868d79116cfa0eefce955557c6fee8c"
+checksum = "dfc6580bb841c5a68e9ef15c77ccc837b40a7504914d52e47b8b0e9bbda25a1d"
 
 [[package]]
 name = "futures-executor"
@@ -987,9 +987,9 @@ dependencies = [
 
 [[package]]
 name = "futures-io"
-version = "0.3.28"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4fff74096e71ed47f8e023204cfd0aa1289cd54ae5430a9523be060cdb849964"
+checksum = "a44623e20b9681a318efdd71c299b6b222ed6f231972bfe2f224ebad6311f0c1"
 
 [[package]]
 name = "futures-lite"
@@ -1008,9 +1008,9 @@ dependencies = [
 
 [[package]]
 name = "futures-macro"
-version = "0.3.28"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "89ca545a94061b6365f2c7355b4b32bd20df3ff95f02da9329b34ccc3bd6ee72"
+checksum = "87750cf4b7a4c0625b1529e4c543c2182106e4dedc60a2a6455e00d212c489ac"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -1019,21 +1019,21 @@ dependencies = [
 
 [[package]]
 name = "futures-sink"
-version = "0.3.29"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e36d3378ee38c2a36ad710c5d30c2911d752cb941c00c72dbabfb786a7970817"
+checksum = "9fb8e00e87438d937621c1c6269e53f536c14d3fbd6a042bb24879e57d474fb5"
 
 [[package]]
 name = "futures-task"
-version = "0.3.28"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "76d3d132be6c0e6aa1534069c705a74a5997a356c0dc2f86a47765e5617c5b65"
+checksum = "38d84fa142264698cdce1a9f9172cf383a0c82de1bddcf3092901442c4097004"
 
 [[package]]
 name = "futures-util"
-version = "0.3.28"
+version = "0.3.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "26b01e40b772d54cf6c6d721c1d1abd0647a0106a12ecaa1c186273392a69533"
+checksum = "3d6401deb83407ab3da39eba7e33987a73c3df0c82b4bb5813ee871c19c41d48"
 dependencies = [
  "futures-core",
  "futures-io",
@@ -1555,9 +1555,9 @@ checksum = "b9e0384b61958566e926dc50660321d12159025e767c18e043daf26b70104c39"
 
 [[package]]
 name = "idna"
-version = "0.3.0"
+version = "0.5.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e14ddfc70884202db2244c223200c204c2bda1bc6e0998d11b5e024d657209e6"
+checksum = "634d9b1461af396cad843f47fdba5597a4f9e6ddd4bfb6ff5d85028c25cb12f6"
 dependencies = [
  "unicode-bidi",
  "unicode-normalization",
@@ -1986,6 +1986,10 @@ checksum = "e4a24736216ec316047a1fc4252e27dabb04218aa4a3f37c6e7ddbf1f9782b54"
 name = "nextchat"
 version = "0.1.0"
 dependencies = [
+ "bytes",
+ "futures-util",
+ "percent-encoding",
+ "reqwest",
  "serde",
  "serde_json",
  "tauri",
@@ -2281,9 +2285,9 @@ checksum = "8835116a5c179084a830efb3adc117ab007512b535bc1a21c991d3b32a6b44dd"
 
 [[package]]
 name = "percent-encoding"
-version = "2.2.0"
+version = "2.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "478c572c3d73181ff3c2539045f6eb99e5491218eae919370993b890cdbdd98e"
+checksum = "e3148f5046208a5d56bcfc03053e3ca6334e51da8dfb19b6cdc8b306fae3283e"
 
 [[package]]
 name = "phf"
@@ -2545,9 +2549,9 @@ checksum = "dc375e1527247fe1a97d8b7156678dfe7c1af2fc075c9a4db3690ecd2a148068"
 
 [[package]]
 name = "proc-macro2"
-version = "1.0.58"
+version = "1.0.86"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fa1fb82fc0c281dd9671101b66b771ebbe1eaf967b96ac8740dcba4b70005ca8"
+checksum = "5e719e8df665df0d1c8fbfd238015744736151d4445ec0836b8e628aae103b77"
 dependencies = [
  "unicode-ident",
 ]
@@ -3889,9 +3893,9 @@ checksum = "1dd624098567895118886609431a7c3b8f516e41d30e0643f03d94592a147e36"
 
 [[package]]
 name = "url"
-version = "2.3.1"
+version = "2.5.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0d68c799ae75762b8c3fe375feb6600ef5602c883c5d21eb51c09f22b83c4643"
+checksum = "22784dbdf76fdde8af1aeda5622b546b422b6fc585325248a2bf9f5e41e94d6c"
 dependencies = [
  "form_urlencoded",
  "idna",

src-tauri/Cargo.toml

@@ -37,6 +37,10 @@ tauri = { version = "1.5.4", features = [ "http-all",
     "window-unminimize",
 ] }
 tauri-plugin-window-state = { git = "https://github.com/tauri-apps/plugins-workspace", branch = "v1" }
+percent-encoding = "2.3.1"
+reqwest = "0.11.18"
+futures-util = "0.3.30"
+bytes = "1.7.2"
 
 [features]
 # this feature is used for production builds or when `devPath` points to the filesystem and the built-in dev server is disabled.

src-tauri/src/main.rs

@@ -1,8 +1,11 @@
 // Prevents additional console window on Windows in release, DO NOT REMOVE!!
 #![cfg_attr(not(debug_assertions), windows_subsystem = "windows")]
 
+mod stream;
+
 fn main() {
   tauri::Builder::default()
+    .invoke_handler(tauri::generate_handler![stream::stream_fetch])
     .plugin(tauri_plugin_window_state::Builder::default().build())
     .run(tauri::generate_context!())
     .expect("error while running tauri application");

src-tauri/src/stream.rs

@@ -0,0 +1,134 @@
+//
+//
+
+use std::time::Duration;
+use std::error::Error;
+use std::sync::atomic::{AtomicU32, Ordering};
+use std::collections::HashMap;
+use futures_util::{StreamExt};
+use reqwest::Client;
+use reqwest::header::{HeaderName, HeaderMap};
+
+static REQUEST_COUNTER: AtomicU32 = AtomicU32::new(0);
+
+#[derive(Debug, Clone, serde::Serialize)]
+pub struct StreamResponse {
+  request_id: u32,
+  status: u16,
+  status_text: String,
+  headers: HashMap<String, String>
+}
+
+#[derive(Clone, serde::Serialize)]
+pub struct EndPayload {
+  request_id: u32,
+  status: u16,
+}
+
+#[derive(Clone, serde::Serialize)]
+pub struct ChunkPayload {
+  request_id: u32,
+  chunk: bytes::Bytes,
+}
+
+#[tauri::command]
+pub async fn stream_fetch(
+  window: tauri::Window,
+  method: String,
+  url: String,
+  headers: HashMap<String, String>,
+  body: Vec<u8>,
+) -> Result<StreamResponse, String> {
+
+  let event_name = "stream-response";
+  let request_id = REQUEST_COUNTER.fetch_add(1, Ordering::SeqCst);
+
+  let mut _headers = HeaderMap::new();
+  for (key, value) in &headers {
+    _headers.insert(key.parse::<HeaderName>().unwrap(), value.parse().unwrap());
+  }
+
+  // println!("method: {:?}", method);
+  // println!("url: {:?}", url);
+  // println!("headers: {:?}", headers);
+  // println!("headers: {:?}", _headers);
+
+  let method = method.parse::<reqwest::Method>().map_err(|err| format!("failed to parse method: {}", err))?;
+  let client = Client::builder()
+    .default_headers(_headers)
+    .redirect(reqwest::redirect::Policy::limited(3))
+    .connect_timeout(Duration::new(3, 0))
+    .build()
+    .map_err(|err| format!("failed to generate client: {}", err))?;
+
+  let mut request = client.request(
+    method.clone(),
+    url.parse::<reqwest::Url>().map_err(|err| format!("failed to parse url: {}", err))?
+  );
+
+  if method == reqwest::Method::POST || method == reqwest::Method::PUT || method == reqwest::Method::PATCH {
+    let body = bytes::Bytes::from(body);
+    // println!("body: {:?}", body);
+    request = request.body(body);
+  }
+
+  // println!("client: {:?}", client);
+  // println!("request: {:?}", request);
+
+  let response_future = request.send();
+
+  let res = response_future.await;
+  let response = match res {
+    Ok(res) => {
+      // get response and emit to client
+      let mut headers = HashMap::new();
+      for (name, value) in res.headers() {
+        headers.insert(
+          name.as_str().to_string(),
+          std::str::from_utf8(value.as_bytes()).unwrap().to_string()
+        );
+      }
+      let status = res.status().as_u16();
+
+      tauri::async_runtime::spawn(async move {
+        let mut stream = res.bytes_stream();
+
+        while let Some(chunk) = stream.next().await {
+          match chunk {
+            Ok(bytes) => {
+              // println!("chunk: {:?}", bytes);
+              if let Err(e) = window.emit(event_name, ChunkPayload{ request_id, chunk: bytes }) {
+                println!("Failed to emit chunk payload: {:?}", e);
+              }
+            }
+            Err(err) => {
+              println!("Error chunk: {:?}", err);
+            }
+          }
+        }
+        if let Err(e) = window.emit(event_name, EndPayload{ request_id, status: 0 }) {
+          println!("Failed to emit end payload: {:?}", e);
+        }
+      });
+
+      StreamResponse {
+        request_id,
+        status,
+        status_text: "OK".to_string(),
+        headers,
+      }
+    }
+    Err(err) => {
+      println!("Error response: {:?}", err.source().expect("REASON").to_string());
+      StreamResponse {
+        request_id,
+        status: 599,
+        status_text: err.source().expect("REASON").to_string(),
+        headers: HashMap::new(),
+      }
+    }
+  };
+  // println!("Response: {:?}", response);
+  Ok(response)
+}
+

src-tauri/tauri.conf.json

@@ -9,7 +9,7 @@
   },
   "package": {
     "productName": "NextChat",
-    "version": "2.15.3"
+    "version": "2.15.4"
   },
   "tauri": {
     "allowlist": {