请求太大的问题[Bug]

[Suaig]

📦 部署方式

Docker

📌 软件版本

2.13.0

💻 系统环境

Windows, Ubuntu

📌 系统版本

Windows 11&Linux lavm-rzea14rft3 5.15.0-105-generic #115-Ubuntu

🌐 浏览器

Chrome, Edge

📌 浏览器版本

Version 126.0.2592.102

🐛 问题描述

GPT4o 面对图片提问的时候，总是Request Entity Too Large。但是同样的图片官方的网页端可以。应该是因为NextChat没有对图片进行大小转换限定之后再发送。

📷 复现步骤

1. 选择GPT4或任何可以识图的模型
2. 上传一张大图片

🚦 期望结果

期望上传大图片的时候会根据模型的最大输入自动转换尺寸。和官网使用体验一样。

📝 补充信息

No response

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Title: The request is too large [Bug]

📦 Deployment method

Docker

📌 Software version

2.13.0

💻 System environment

Windows, Ubuntu

📌 System version

Windows 11&Linux lavm-rzea14rft3 5.15.0-105-generic #115-Ubuntu

🌐 Browser

Chrome, Edge

📌 Browser version

Version 126.0.2592.102

🐛 Problem description

GPT4o always requests Entity Too Large when asked about pictures. But the same picture can be found on the official web page. It may be because NextChat does not limit the size of the image before sending it.

📷 Steps to reproduce

1. Select GPT4 or any model that can recognize images
2. Upload a large picture

🚦 Expected results

It is expected that when uploading large images, the size will be automatically converted according to the maximum input of the model. The experience is the same as the official website.

📝 Supplementary information

No response

[Dogtiti]

下个版本会优化图片存储

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

The next version will optimize image storage

[lloydzhou]

Request Entity Too Large

应该是因为当前项目是纯前端项目，发送图片到api的时候，使用的是base64的url导致请求体过大。

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Request Entity Too Large

It should be because the current project is a pure front-end project. When sending pictures to the API, the base64 URL is used, causing the request body to be too large.