Getting Started

1. Get Your API Key

1. Sign up: Click Login in the top right and sign up using single sign-on
2. Create API Key: Go to API Keys and click Create API Key
3. Generate: Enter a name, set expiration, and click Generate Key

2. Explore Available Models

Want to see what other models are available? You can list all models using:

Code
 
curl -X GET https://api.publicai.co/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "User-Agent: MyApp/1.0"

Or try it interactively:

Open in Playground

3. Making Your First Request

Once you have an API key, you can start making requests to our AI models. Here's a simple example:

Code
 
curl -X POST https://api.publicai.co/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "User-Agent: MyApp/1.0" \
  -d '{
    "model": "swiss-ai/apertus-8b-instruct",
    "messages": [
      {
        "role": "user",
        "content": "Hello! Can you help me understand open-source AI?"
      }
    ]
  }'

Or try it interactively:

Open in Playground

Authentication

All API requests require:

1. API Key: Include your key in the Authorization header:

   Code
    
   Authorization: Bearer YOUR_API_KEY

2. User-Agent Header: Specify a User-Agent to prevent bot spamming:

   Code
    
   User-Agent: MyApp/1.0

Model Configuration

Our recommended inference parameters, which we use on our consumer frontend at publicai.co:

• Context Window: 65,536 tokens
• Max Output Tokens: 8,192 tokens
• Temperature: 0.8 (recommended by swiss-ai)
• Top-p: 0.9 (recommended by swiss-ai)

Additional parameter tuning and documentation will be available in future updates.

Rate Limits

To ensure fair usage across all users:

• During Swiss AI Weeks: 20 requests per minute

Support

• Documentation: Explore the full API Reference
• Community: Join our discussions at github.com/forpublicai
• Issues: Report problems via GitHub Issues

What's Next?

• Explore the complete API Reference for detailed endpoint documentation
• Check out example implementations in our GitHub repositories
• Join our community to stay updated on new model releases