Skip to content

Using Hydra in Lobe Chat

Lobe Chat is an open-source modern design AI chat framework.
It supports multiple AI providers (OpenAI/Claude 3/Gemini/Ollama/Qwen/DeepSeek), knowledge bases (file upload/knowledge management/RAG), multimodal features (vision/TTS/plugins/artwork), and offers one-click free deployment of your private ChatGPT/Claude applications.

Lobe Chat

Install Lobe Chat

You can refer to the Lobe Chat official documentation to download and install it.
Lobe Chat offers multiple deployment and installation methods.

This document uses Docker as an example to explain how to use Hydra’s model services.

# Lobe Chat supports configuring API Key and API Host directly at deployment
$ docker run -d -p 3210:3210 \
    -e OPENAI_API_KEY=sk-xxxx \  # Enter your API Key
    -e OPENAI_PROXY_URL=https://chat.d.run/v1 \  # Enter your API Host
    -e ENABLED_OLLAMA=0 \
    -e ACCESS_CODE=drun \
    --name lobe-chat \
    lobehub/lobe-chat:latest

Configure Lobe Chat

Lobe Chat also supports adding model provider configurations after deployment.

Lobe Chat

Add your API Key and API Host obtained from Hydra.

  • API Key: Enter your API Key

  • API Host:

    • For MaaS, use https://chat.d.run
    • For independently deployed model services, check model instance details, usually https://<region>.d.run

  • Configure custom model, e.g., public/deepseek-r1

Lobe Chat Usage Demo

Lobe Chat

Comments