DuckDuckGo AI

中文说明

Deploy a free DuckDuckGo AI using Cloudflare Worker, supporting models like gpt-3.5-turbo-0125,claude-3-haiku-20240307, ...

DuckDuckGo AI: https://duckduckgo.com/?q=DuckDuckGo&ia=chat

Deployment

git clone https://github.com/anhao/duckduckgo-ai.git
cd duckduckgo-ai
npm install
npm run deploy

Supported Models

• gpt-3.5-turbo-0125
• claude-3-haiku-20240307
• meta-llama/Llama-3-70b-chat-hf
• mistralai/Mixtral-8x7B-Instruct-v0.1

Usage

Replace worker_url with your own:

curl https://worker_url/v1/chat/completions \
-H "Authorization: Bearer $YOU_APIKEY" \
-H "Content-Type: application/json" \
-d '{
    "model": "gpt-3.5-turbo-0125",
    "messages": [{"role": "user", "content": "Hello"}],
}'

Authorization Access

In the wrangler.toml file, set the value of the apikey parameter. If not set, anyone can make requests.

[vars]
apikey = "" ## Set the apikey value, if not set, anyone can make requests.

Known Issues

Currently, it does not support continuous dialogue, meaning the messages can only be passed once.

If you want to pass multiple messages, you need to manually pass the x-vqd-4 parameter value to the request header.

The x-vqd-4 parameter will be returned in the response header after the first dialogue, and subsequent requests need to pass the x-vqd-4 value. A new x-vqd-4 parameter is required for each new dialogue added.

• [ ] Automatically pass the x-vqd-4 by storing the conversation hash in Cloudflare KV