Skip to main content

​
Prerequisites

Before you begin, make sure you have:
  • A Meibel AI account (sign up at console.meibel.ai if you don’t have one)
  • An API key (generated from the Console under Settings > API Keys)
  • Python 3.9+ (if using the Python SDK)
  • Basic knowledge of API requests

​
Step 1: Set Up Authentication

All requests to the Meibel AI API require authentication using your API key. You’ll need to include this key in the Authorization header of your requests. 
# Store your API key in an environment variable (recommended)
export MEIBEL_API_KEY="your_api_key_here"

# Make a test request to list datasources
curl -X GET "https://api.meibel.ai/v1/datasources" \
  -H "Authorization: Bearer $MEIBEL_API_KEY" \
  -H "Content-Type: application/json"

​
Step 2: Create a Datasource

Datasources are containers for your data elements. Let’s create a datasource for our project: 
curl -X POST "https://api.meibel.ai/v1/datasources" \
  -H "Authorization: Bearer $MEIBEL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Product Documentation",
    "description": "Documentation for our product features and APIs"
  }'

​
Step 3: Add a Data Element

Now, let’s add some content to our datasource: 
# Replace YOUR_DATASOURCE_ID with the ID from the previous response
curl -X POST "https://api.meibel.ai/v1/datasources/YOUR_DATASOURCE_ID/dataelements" \
  -H "Authorization: Bearer $MEIBEL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "API Overview",
    "content": "The Meibel AI API provides several endpoints for managing data and running AI completions. The main endpoints include /datasources, /completions, and /experiences.",
    "metadata": {
      "category": "documentation",
      "section": "api"
    }
  }'

​
Step 4: Run a Completion

Now, let’s use our datasource to power a completion with context retrieval: 
curl -X POST "https://api.meibel.ai/v1/completions" \
  -H "Authorization: Bearer $MEIBEL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "What are the main endpoints in your API?"}
    ],
    "datasource_ids": ["YOUR_DATASOURCE_ID"],
    "confidence_threshold": 0.7,
    "execution_control": {
      "enable_tracing": true
    }
  }'

​
Step 5: Stream a Response

For longer responses or real-time interactions, you can use streaming: 
# Stream a completion
stream = client.rag.chat_stream(
    messages=[
        {"role": "user", "content": "Explain how the Meibel AI confidence scoring works"}
    ],
    datasource_ids=[datasource_id],
    confidence_threshold=0.7
)

# Process the stream
for chunk in stream:
    if chunk.choices and len(chunk.choices) > 0 and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

​
Next Steps

Congratulations! You’ve successfully made your first requests to the Meibel AI API. Here are some next steps to explore:

Explore Confidence Scoring

Learn how to leverage our confidence scoring system for higher accuracy

Set Up Execution Control

Gain precise control over AI execution flow and traceability

Configure Webhooks

Set up webhooks for asynchronous notifications

Browse the API Reference

Explore all available endpoints and parameters
Remember to secure your API keys and follow best practices for rate limiting and error handling in your production applications.