Skip to main content

API Keys

API keys are required to access Enclava's public API endpoints.

Creating an API Key​

Via Web Interface​

  1. Navigate to API Keys in the sidebar

  2. Click Create API Key

  3. Fill in the configuration form:

    Basic Settings:

    • Name: Descriptive name (e.g., "Production App", "Dev Bot")
    • Description: Optional additional information
    • Prefix: Auto-generated prefix (default: en_)

    Model Access:

    • Allowed Models: Leave empty for all models, or specify (e.g., gpt-4,gpt-3.5-turbo)
    • Purpose: Restrict key to specific use cases

    Budget Limits:

    • Daily Spend Limit: Maximum spend per day (e.g., 10.00)
    • Monthly Spend Limit: Maximum spend per month (e.g., 300.00)
    • Daily Token Limit: Maximum tokens per day (e.g., 100000)
    • Alert Threshold: Percentage at which to send alerts (e.g., 0.8 for 80%)

  4. Click Create to generate the key

Via API​

import requests

response = requests.post(
    "http://localhost/api-internal/v1/api-keys",
    headers={
        "Authorization": "Bearer YOUR_JWT_TOKEN"
    },
    json={
        "name": "Production App",
        "description": "API key for production application",
        "allowed_models": ["gpt-4", "gpt-3.5-turbo"],
        "daily_spend_limit": 50.00,
        "monthly_spend_limit": 1500.00,
        "daily_token_limit": 500000
    }
)

api_key = response.json()["api_key"]
print(f"Your API key: {api_key}")

⚠️ Important: Copy the API key immediately - you won't see it again!

API Key Format​

Enclava API keys follow this format:

en_<random-string>

Example: en_8f3d4e2a9b1c6d5f7e8a3b4c9d2f1e6a

The prefix en_ identifies Enclava keys and helps with organization.

Naming Best Practices​

Use descriptive names to identify keys easily:

PurposeExample Name
Development environmentdev-app-v1
Production web appprod-web-app
Chatbot widgetchatbot-widget-frontend
Internal toolinternal-data-processor
Testingtest-script-automation

Setting Model Access​

All Models (Default)​

Leave "Allowed Models" empty:

{
  "name": "Full Access Key",
  "allowed_models": []
}

This key can access all models available on your platform.

Specific Models​

Specify models explicitly:

{
  "name": "GPT-4 Only Key",
  "allowed_models": ["gpt-4"]
}

This key can only use gpt-4, useful for cost control.

Multiple Models​

Specify comma-separated list:

{
  "name": "Production Models",
  "allowed_models": ["gpt-4", "gpt-3.5-turbo"]
}

Setting Budget Limits​

Daily Spend Limit​

Prevents unexpected costs on a daily basis:

{
  "daily_spend_limit": 25.00
}

After $25 of usage in a day, requests using this key will be rejected.

Monthly Spend Limit​

Controls overall monthly spending:

{
  "monthly_spend_limit": 750.00
}

Good for predictable monthly budgets.

Token Limits​

Control usage by token count:

{
  "daily_token_limit": 250000
}

Useful when costs vary by model (e.g., cheaper models can use more tokens).

Alert Threshold​

Get notified before limits are reached:

{
  "daily_spend_limit": 100.00,
  "alert_threshold": 0.8
}

You'll be notified when you reach 80% of your daily limit.

Testing Your New Key​

Verify Key Works​

import requests

response = requests.get(
    "http://localhost/api/v1/models",
    headers={"Authorization": "Bearer en_xxxxxxxxxxxxxxx"}
)

if response.status_code == 200:
    print("✓ API key is working")
    models = response.json()["data"]
    print(f"Available models: {[m['id'] for m in models]}")
else:
    print(f"✗ Error: {response.status_code}")
    print(response.json())

Test Model Access​

response = requests.post(
    "http://localhost/api/v1/chat/completions",
    headers={"Authorization": "Bearer en_xxxxxxxxxxxxxxx"},
    json={
        "model": "gpt-4",
        "messages": [{"role": "user", "content": "Test"}]
    }
)

print(response.json()["choices"][0]["message"]["content"])

Common Creation Issues​

Key Not Generated​

Problem: No API key shown after creation

Solution:

  • Refresh the page and check API Keys list
  • Keys are only shown once during creation
  • If missed, delete and create a new key

Budget Too Low​

Problem: Requests fail immediately

Solution:

  • Verify budget limits are reasonable
  • Start with higher limits for testing
  • Monitor usage to set appropriate limits

Model Access Error​

Problem: "Model not allowed" error

Solution:

  • Check allowed_models list
  • Verify model name matches exactly (case-sensitive)
  • Try without model restrictions to test

Next Steps​

After creating your API key: