概要

To further enhance the development experience and simplify interface integration, AI Gateway now introduces API V2 version. This version has multiple improvements and optimizations, especially in the unification of the request body and response data format.
Major Updates: 
Unified request interface address to avoid frequent switching of interface addresses by developers.
Unified Request Body Format: All API requests now use a consistent JSON format, simplifying the data construction process for clients. 
Standardized Response Data Structure: A unified response format helps developers quickly understand and process returned information. It includes status codes, messages, and data body, making error handling and success feedback more intuitive and clear.
Enhanced Error Handling Mechanism: Clear error code definitions facilitate problem localization and resolution. Detailed error descriptions help developers quickly take corrective actions.
﻿
We believe these improvements will greatly enhance your development efficiency and user experience of the application. Thank you for your continued support and trust. If you have any questions or feedback, please feel free to contact our technical support team (you can reach us via the feedback button in the lower right corner of the page).
﻿
Adapted LLM Interfaces
This document aims to provide developers with a detailed guide on using AI Gateway V2 to integrate various large language model (LLM) interfaces.
The currently adapted LLM interfaces are listed below
LLM 
API version
Document Link
API address
Supported Inference Access Points
Baidu-QianFan
Dialogue chat v1
Click to access
https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/${model}?access_token=${apiKey}
Refer to https://cloud.baidu.com/doc/WENXINWORKSHOP/s/nluv1jxlp left navigation
Baidu-QianFan
Dialogue chat v2
Click to access
https://qianfan.baidubce.com/v2/chat/completions
Refer to 
https://cloud.baidu.com/doc/WENXINWORKSHOP/s/Fm2vrveyu#%E6%94%AF%E6%8C%81%E6%A8%A1%E5%9E%8B%E5%88%97%E8%A1%A8
Byte-Doubao
ChatCompletions - Text Dialogue
Click to access
https://ark.cn-beijing.volces.com/api/v3/chat/completions
Refer to https://www.volcengine.com/docs/82379/1099522
Tencent-HunYuan
HunYuan Text Generation
Click to access
https://hunyuan.tencentcloudapi.com
Optional values include hunyuan-lite, hunyuan-standard, hunyuan-standard-256K, hunyuan-pro, hunyuan-code, hunyuan-role, hunyuan-functioncall, hunyuan-vision, hunyuan-turbo, hunyuan-turbo-latest, hunyuan-large
Moonshot-KIMI
chat
Click to access
https://api.moonshot.cn/v1/chat/completions
Optional values include moonshot-v1-8k, moonshot-v1-32k, moonshot-v1-128k
MINIMAX
v2
Click to access
https://api.minimax.chat/v1/text/chatcompletion_v2
Optional values include abab7-chat-preview, abab6.5s-chat, abab6.5t-chat, abab6.5g-chat, abab5.5-chat, abab5.5s-chat
Alibaba Tongyi Qianwen
OpenAI Chat
Click to access
https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions
Refer to https://help.aliyun.com/zh/dashscope/developer-reference/compatibility-of-openai-with-dashscope?spm=a2c4g.11186623.0.i6
OPENAI
Dialogue chat
Click to access
https://api.openai.com/v1/chat/completions
Please refer to the official website instructions
GEMINI
Generate text based on plain text input
Click to access
https://generativelanguage.googleapis.com/v1beta/models/gemini-1.5-flash:${body?.stream? 'streamGenerateContent?alt=sse' : 'generateContent'}?key=${X-Goog_Api_Key}
Please refer to the official website instructions
﻿
﻿

Major Updates:
Adapted LLM Interfaces

LLM	API version	Document Link	API address	Supported Inference Access Points
Baidu-QianFan	Dialogue chat v1	Click to access	https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/${model}?access_token=${apiKey}	Refer to https://cloud.baidu.com/doc/WENXINWORKSHOP/s/nluv1jxlp left navigation
Baidu-QianFan	Dialogue chat v2	Click to access	https://qianfan.baidubce.com/v2/chat/completions	Refer to https://cloud.baidu.com/doc/WENXINWORKSHOP/s/Fm2vrveyu#%E6%94%AF%E6%8C%81%E6%A8%A1%E5%9E%8B%E5%88%97%E8%A1%A8
Byte-Doubao	ChatCompletions - Text Dialogue	Click to access	https://ark.cn-beijing.volces.com/api/v3/chat/completions	Refer to https://www.volcengine.com/docs/82379/1099522
Tencent-HunYuan	HunYuan Text Generation	Click to access	https://hunyuan.tencentcloudapi.com	Optional values include hunyuan-lite, hunyuan-standard, hunyuan-standard-256K, hunyuan-pro, hunyuan-code, hunyuan-role, hunyuan-functioncall, hunyuan-vision, hunyuan-turbo, hunyuan-turbo-latest, hunyuan-large
Moonshot-KIMI	chat	Click to access	https://api.moonshot.cn/v1/chat/completions	Optional values include moonshot-v1-8k, moonshot-v1-32k, moonshot-v1-128k
MINIMAX	v2	Click to access	https://api.minimax.chat/v1/text/chatcompletion_v2	Optional values include abab7-chat-preview, abab6.5s-chat, abab6.5t-chat, abab6.5g-chat, abab5.5-chat, abab5.5s-chat
Alibaba Tongyi Qianwen	OpenAI Chat	Click to access	https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions	Refer to https://help.aliyun.com/zh/dashscope/developer-reference/compatibility-of-openai-with-dashscope?spm=a2c4g.11186623.0.i6
OPENAI	Dialogue chat	Click to access	https://api.openai.com/v1/chat/completions	Please refer to the official website instructions
GEMINI	Generate text based on plain text input	Click to access	https://generativelanguage.googleapis.com/v1beta/models/gemini-1.5-flash:${body?.stream? 'streamGenerateContent?alt=sse' : 'generateContent'}?key=${X-Goog_Api_Key}	Please refer to the official website instructions

概要

Major Updates:

Adapted LLM Interfaces

﻿