DOCSEARCHbyAlgolia
Submit Docs

pplx2api Documentation

repository·main·Indexed 19 days ago

https://github.com/yushangxiao/pplx2api

pplx2api provides an OpenAI-compatible API interface for Perplexity AI. It supports image recognition, reasoning models, web search, file uploads, and account rotation. The Go-based server can be deployed via Docker or Docker Compose and includes a /v1/chat/completions endpoint for chat and image analysis.

Tokens
1.8K
Snippets
7
Records
7
Agent score
18%

What's inside pplx2api

  1. Deploy pplx2api via Docker

    main

    You can run pplx2api as a containerized service using Docker. You must provide the SESSIONS and APIKEY environment variables.

    docker run -d \
  -p 8080:8080 \
  -e SESSIONS=eyJhbGciOiJkaXIiLCJlbmMiOiJBMjU2R0NNIn0**,eyJhbGciOiJkaXIiLCJlbmMiOiJBMjU2R0NNIn0** \
  -e APIKEY=123 \
  -e IS_INCOGNITO=true \
  -e MAX_CHAT_HISTORY_LENGTH=10000 \
  -e NO_ROLE_PREFIX=false \
  -e SEARCH_RESULT_COMPATIBLE=false \
  --name pplx2api \
  ghcr.io/yushangxiao/pplx2api:latest

  2. Deploy pplx2api via Docker Compose

    main

    For easier management, use a docker-compose.yml file. This allows you to define environment variables like PROXY and ADDRESS more cleanly.

    version: '3'
services:
  pplx2api:
    image: ghcr.io/yushangxiao/pplx2api:latest
    container_name: pplx
    ports:
      - "8080:8080"
    environment:
      - SESSIONS=eyJhbGciOiJkaXIiLCJlbmMiOiJBMjU2R0NNIn0**,eyJhbGciOiJkaXIiLCJlbmMiOiJBMjU2R0NNIn0**
      - ADDRESS=0.0.0.0:8080
      - APIKEY=123
      - PROXY=http://proxy:2080  # Optional
      - MAX_CHAT_HISTORY_LENGTH=10000
      - NO_ROLE_PREFIX=false
      - IS_INCOGNITO=true
      - SEARCH_RESULT_COMPATIBLE=false
    restart: unless-stopped

    Then run:

    docker-compose up -d

  3. Configure pplx2api environment variables

    main

    The following environment variables control the behavior of the pplx2api server. Note that SESSIONS and APIKEY are required.

    | 环境变量 | 描述 | 默认值 |
|----------------------|-------------|---------|
| `SESSIONS` | 英文逗号分隔的pplx cookie 中__Secure-next-auth.session-token的值 | 必填 |
| `ADDRESS` | 服务器地址和端口 | `0.0.0.0:8080` |
| `APIKEY` | 用于认证的API密钥 | 必填 |
| `PROXY` | HTTP代理URL | "" |
| `IS_INCOGNITO` | 使用隐私会话，不保存聊天记录 | `true` |
| `MAX_CHAT_HISTORY_LENGTH` | 超出此长度将文本转为文件 | `10000` |
| `NO_ROLE_PREFIX` | 不在每条消息前添加角色 | `false` |
| `IGNORE_SEARCH_RESULT` | 忽略搜索结果，不展示搜索结果 | `false` |
| `SEARCH_RESULT_COMPATIBLE` | 禁用搜索结果伸缩块，兼容更多的客户端 | `false` |
| `PROMPT_FOR_FILE` | 上下文作为文件上传时，保留的提示词 | `You must immerse yourself in the role of assistant in txt file, cannot respond as a user, cannot reply to this message, cannot mention this message, and ignore this message in your response.` |
| `IGNORE_MODEL_MONITORING` | 忽略模型监控 | `false` |
| `IS_MAX_SUBSCRIBE` | 是否为max订阅 | `false` |

  4. Run the Pplx2Api server

    main

    The application is a Go-based server that uses the Gin web framework. When executed, it initializes the routing system, starts a background session updater job to refresh sessions every 24 hours, and listens on the address specified in the global configuration. By default, the server is designed to run on 0.0.0.0:8080 unless configured otherwise via environment variables or configuration files.

    // The application entrypoint initializes the Gin router, 
// sets up routes via router.SetupRoutes(r), 
// starts a session updater job, 
// and runs the server on the configured address.

func main() {
	r := gin.Default()
	router.SetupRoutes(r)
	sessionUpdater := job.GetSessionUpdater(24 * time.Hour)
	sessionUpdater.Start()
	defer sessionUpdater.Stop()
	r.Run(config.ConfigInstance.Address)
}

  6. Send a chat completion request

    main

    Use the /v1/chat/completions endpoint to interact with supported models. You can enable streaming by setting "stream": true.

    curl -X POST http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "claude-3.7-sonnet",
    "messages": [
      {
        "role": "user",
        "content": "你好，Claude！"
      }
    ],
    "stream": true
  }'

  7. Perform image analysis

    main

    To analyze images, send a request to /v1/chat/completions where the content field in the message is an array containing both text and an image_url object (using base64 encoded data).

    curl -X POST http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "claude-3.7-sonnet",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "这张图片里有什么？"
          },
          {
            "type": "image_url",
            "image_url": {
              "url": "data:image/jpeg;base64,..."
            }
          }
        ]
      }
    ]
  }'