The main updates of GeekAI-PLUS-v4.2.2 include support for the official DeepSeek inference model API, which previously only supported relay API access. Additionally, the GPT-4O image generation model has been launched.

Update Content

  • Feature Optimization: When enabling the graphical CAPTCHA feature, the system now checks whether the API service is configured to prevent login issues after enabling.
  • Feature Optimization: Support for the native DeepSeek inference model API. The chat API KEY now supports setting a complete API path, such as https://api.aiggr.com/v1/chat/completions.
  • Feature Optimization: Support for GPT-4o image editing functionality.
  • New Feature: The chat page now supports AI voice output (TTS).
  • Feature Optimization: Replaced the waterfall flow component to improve user experience.
  • Feature Optimization: Automatic caching of the last result when generating mind maps.
  • Feature Optimization: Optimized the MJ drawing page and added support for the MJ-V7 model.
  • Feature Optimization: Added a one-click login link generation feature in the backend management.

1. Update Database

ALTER TABLE `chatgpt_chat_models` ADD `options` TEXT NOT NULL COMMENT '模型自定义选项' AFTER `key_id`;

2. Update Image Version

Edit docker-compose.yaml and update the image versions of geekai-api and geekai-web to v4.2.2.

# Backend API
geekai-api:
  image: registry.cn-shenzhen.aliyuncs.com/geekmaster/geekai-plus-api:v4.2.2-amd64
  container_name: geekai-api
  restart: always
  ports:
    - '5678:5678'
    - '6012:6012'
  volumes:
    - /usr/share/zoneinfo/Asia/Shanghai:/etc/localtime
    - ./conf/config.toml:/var/www/app/config.toml
    - ./logs:/var/www/app/logs
    - ./static:/var/www/app/static
    - ./res/sd:/var/www/app/res/sd
    - ./certs:/var/www/app/certs
    - ./data:/var/www/app/data
  environment:
    - APP_DEBUG=false
    - LOG_LEVEL=info
    - CONFIG_FILE=config.toml

# Frontend Application
geekai-web:
  image: registry.cn-shenzhen.aliyuncs.com/geekmaster/geekai-plus-web:v4.2.2-amd64
  container_name: geekai-web
  restart: always
  ports:
    - '9004:8080'
  volumes:
    - ./logs/nginx:/var/log/nginx
    - ./conf/nginx/conf.d:/etc/nginx/conf.d
    - ./conf/nginx/nginx.conf:/etc/nginx/nginx.conf
    - ./ssl:/etc/nginx/ssl

3. Restart Containers

docker-compose down
docker-compose up -d

4. Feature Configuration

1. Add TTS Model

When adding a model, you can now add a TTS model for voice output on the chat page.

Add TTS Model

After adding, users can select which TTS model to use on the chat page.

Select TTS Model

TTS Model Effect

2. Support for Complete API Path

When adding an API KEY, you only need to fill in the BASE URL by default. The system will automatically access the path $BASE_URL/v1/chat/completions. However, some users may use third-party model APIs from providers like Alibaba Cloud or China Mobile, where the API path is not the standard /v1/chat/completions, resulting in 404 errors. In such cases, you need to fill in the complete API path.

Complete API Path

3. Generate One-Time Login Link

If you want to let users experience the system without registering, you can add an account for them in the backend and generate a one-time login link. Users can click this link to directly log in and try the system.

One-Time Login Link