Skip to main content
Track, measure, and optimize your API usage like a pro Ever wondered which of your apps is burning through your API budget? Or need to bill customers accurately for their AI usage? App Attribution is your answer – a powerful feature that lets you tag, track, and analyze every API call with surgical precision.

​
What’s in it for you?

📊 Crystal-clear usage insights – See exactly how each app, project, or customer uses your API
💰 Bulletproof cost allocation – No more guesswork when dividing bills
🚀 Performance monitoring – Track response times and success rates per app
📈 Data-driven decisions – Export detailed reports and spot trends instantly
🛡️ Smart spending controls – Set limits and get alerts before things get expensive

​
Getting Started: Two Ways to Tag Your Requests

Clean, simple, and works everywhere 
import requests

url = "https://api.anyapi.ai/api/v1/chat/completions"
headers = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json",
    
    # 🏷️ Your attribution tags
    "X-AnyAPI-App-ID": "my-chatbot-v2",           # Required: Your app's unique ID
    "X-AnyAPI-App-Name": "Customer Support Bot",   # Optional: Human-readable name
    "X-AnyAPI-App-Version": "2.1.0",              # Optional: Track versions
    "X-AnyAPI-User-ID": "user_12345",             # Optional: End user tracking
    "X-AnyAPI-Session-ID": "session_abc123"       # Optional: Session tracking
}

payload = {
    "model": "openai/gpt-4o",
    "messages": [{"role": "user", "content": "Hello, how can you help me?"}]
}

response = requests.post(url, headers=headers, json=payload)

​
Method 2: Request Body Metadata

Perfect for complex attribution scenarios 
payload = {
    "model": "anthropic/claude-3-5-sonnet-20241022",
    "messages": [{"role": "user", "content": "Analyze this quarterly data..."}],
    
    # 🎯 Rich metadata for detailed tracking
    "metadata": {
        "app_id": "data-analyzer",
        "app_name": "Business Intelligence Tool", 
        "app_version": "1.5.2",
        "user_id": "analyst_jane_doe",
        "project": "q4-sales-analysis",
        "department": "marketing",
        "region": "us-west"
    }
}

​
Attribution Fields Reference

​
Must-Have

  • app_id – Your app’s unique identifier (this is the only required field)

​
Nice-to-Have (But Really Useful)

  • app_name – Human-friendly app name for dashboards
  • app_version – Track performance across versions
  • user_id – See per-user usage patterns
  • session_id – Follow conversation threads
  • project – Organize by campaigns or initiatives
  • environment – Separate prod from staging costs
  • region – Geographic usage distribution

​
Real-World Implementation

​
Single App Setup

Perfect for most developers 
class AnyAPIClient:
    def __init__(self, api_key, app_id, app_name=None, app_version=None):
        self.api_key = api_key
        self.app_id = app_id
        self.app_name = app_name
        self.app_version = app_version
        self.base_url = "https://api.anyapi.ai/api/v1"
    
    def _get_headers(self, user_id=None, session_id=None):
        """Smart header generation with attribution"""
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json",
            "X-AnyAPI-App-ID": self.app_id
        }
        
        # Add optional fields only when they exist
        if self.app_name:
            headers["X-AnyAPI-App-Name"] = self.app_name
        if self.app_version:
            headers["X-AnyAPI-App-Version"] = self.app_version
        if user_id:
            headers["X-AnyAPI-User-ID"] = user_id
        if session_id:
            headers["X-AnyAPI-Session-ID"] = session_id
        
        return headers
    
    def chat_completion(self, messages, model="openai/gpt-4o", 
                       user_id=None, session_id=None, **kwargs):
        """Make attributed requests effortlessly"""
        
        headers = self._get_headers(user_id, session_id)
        payload = {"model": model, "messages": messages, **kwargs}
        
        response = requests.post(
            f"{self.base_url}/chat/completions",
            headers=headers,
            json=payload
        )
        return response.json()

# 🚀 Usage is dead simple
client = AnyAPIClient(
    api_key=API_KEY,
    app_id="customer-support-bot",
    app_name="Customer Support Assistant", 
    app_version="3.2.1"
)

# Every request gets properly attributed automatically
result = client.chat_completion(
    messages=[{"role": "user", "content": "How do I reset my password?"}],
    user_id="customer_54321",
    session_id="support_session_789"
)

​
Multi-Tenant SaaS Setup

When you’re serving multiple customers 
class MultiTenantAPIClient:
    def __init__(self, api_key, base_app_id):
        self.api_key = api_key
        self.base_app_id = base_app_id
        self.base_url = "https://api.anyapi.ai/api/v1"
    
    def make_request_for_tenant(self, tenant_id, messages, 
                               model="openai/gpt-4o", **kwargs):
        """Tenant-specific attribution made easy"""
        
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json",
            "X-AnyAPI-App-ID": f"{self.base_app_id}-{tenant_id}",
            "X-AnyAPI-App-Name": f"SaaS App - {tenant_id}",
            "X-AnyAPI-User-ID": kwargs.get("user_id"),
            "X-AnyAPI-Tenant-ID": tenant_id
        }
        
        # Rich metadata for business intelligence
        payload = {
            "model": model,
            "messages": messages,
            "metadata": {
                "tenant_id": tenant_id,
                "subscription_tier": kwargs.get("tier", "standard"),
                "feature_flags": kwargs.get("features", []),
                "billing_cycle": kwargs.get("billing_cycle", "monthly")
            }
        }
        
        response = requests.post(f"{self.base_url}/chat/completions",
                               headers=headers, json=payload)
        return response.json()

# 💼 Perfect for SaaS platforms
client = MultiTenantAPIClient(API_KEY, "my-saas-platform")

# Company A gets premium features
result_a = client.make_request_for_tenant(
    tenant_id="company-abc",
    messages=[{"role": "user", "content": "Generate quarterly projections..."}],
    user_id="ceo_alice",
    tier="enterprise",
    features=["advanced_analytics", "custom_models"]
)

# Startup B gets the essentials  
result_b = client.make_request_for_tenant(
    tenant_id="startup-xyz", 
    messages=[{"role": "user", "content": "Help brainstorm marketing ideas..."}],
    user_id="founder_bob",
    tier="startup"
)

​
Analytics That Actually Matter

​
Get Usage Insights

See where your money’s going 
def get_usage_by_app(app_id, start_date, end_date):
    """Pull detailed usage stats for any app"""
    
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    
    params = {
        "app_id": app_id,
        "start_date": start_date,
        "end_date": end_date,
        "group_by": ["date", "model", "user_id"]  # Slice and dice your data
    }
    
    response = requests.get(
        "https://api.anyapi.ai/api/v1/analytics/usage",
        headers=headers,
        params=params
    )
    
    return response.json()

# 📊 Get the last 30 days of data
import datetime

end_date = datetime.datetime.now()
start_date = end_date - datetime.timedelta(days=30)

usage_data = get_usage_by_app(
    "customer-support-bot",
    start_date.isoformat(),
    end_date.isoformat()
)

print(f"🔥 Total requests: {usage_data['total_requests']:,}")
print(f"💰 Total cost: ${usage_data['total_cost']:.2f}")
print(f"⚡ Average response time: {usage_data['avg_response_time']}ms")

​
Cost Allocation Reports

Perfect for finance teams and client billing 
def generate_cost_allocation_report(start_date, end_date):
    """Create executive-ready cost breakdowns"""
    
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    
    payload = {
        "start_date": start_date,
        "end_date": end_date,
        "group_by": ["app_id", "model"],
        "metrics": ["requests", "cost", "tokens", "avg_response_time"]
    }
    
    response = requests.post(
        "https://api.anyapi.ai/api/v1/analytics/cost-allocation",
        headers=headers,
        json=payload
    )
    
    data = response.json()
    
    # 📋 Beautiful console report
    print("💎 Cost Allocation Report")
    print("=" * 50)
    
    for app in data["apps"]:
        print(f"\n🚀 App: {app['app_name']} ({app['app_id']})")
        print(f"   📊 Requests: {app['total_requests']:,}")
        print(f"   💵 Cost: ${app['total_cost']:.2f}")
        print(f"   ⚡ Avg Response: {app['avg_response_time']}ms")
        
        for model in app["models"]:
            print(f"      🤖 {model['model']}: {model['requests']:,} requests, ${model['cost']:.2f}")
    
    return data

​
Smart Usage Controls

​
Set Spending Limits

Sleep better knowing your costs are under control 
def set_app_usage_limit(app_id, monthly_limit_dollars):
    """Set bulletproof spending limits"""
    
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    
    payload = {
        "app_id": app_id,
        "limit_type": "monthly",
        "limit_amount": monthly_limit_dollars * 100,  # Convert to credits
        "alert_thresholds": [50, 80, 95]  # Get warned before it's too late
    }
    
    response = requests.post(
        "https://api.anyapi.ai/api/v1/limits/app",
        headers=headers,
        json=payload
    )
    
    return response.json()

# 🛡️ Set a $100 monthly limit 
limit_result = set_app_usage_limit("customer-support-bot", 100)
print(f"✅ Limit set successfully: {limit_result['success']}")

​
Real-Time Usage Monitoring

Stay on top of your spending 
def check_app_usage_status(app_id):
    """Get instant usage updates"""
    
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    }
    
    response = requests.get(
        f"https://api.anyapi.ai/api/v1/usage/app/{app_id}/current",
        headers=headers
    )
    
    data = response.json()
    usage_percent = (data["current_usage"] / data["limit"]) * 100
    
    print(f"📱 App: {data['app_name']}")
    print(f"💰 Current: ${data['current_usage']/100:.2f}")
    print(f"🎯 Limit: ${data['limit']/100:.2f}")
    print(f"📊 Usage: {usage_percent:.1f}%")
    
    # Smart alerts
    if usage_percent > 90:
        print("🚨 ALERT: You're almost at your limit!")
    elif usage_percent > 75:
        print("⚠️  HEADS UP: Usage is getting high")
    else:
        print("✅ All good! Spending is under control")
    
    return data

​
Pro Tips for Success

​
🎯 Smart App IDs

  • Be descriptive: Use customer-chat-prod not app1
  • Stay consistent: Pick a naming pattern and stick to it
  • Include environment: Separate prod, staging, dev

​
📊 Meaningful Data

  • Track what matters: User IDs for SaaS, session IDs for chat apps
  • Version everything: Catch performance regressions early
  • Add context: Project names, feature flags, A/B test groups

​
💰 Cost Management

  • Set realistic limits: Start conservative, adjust as needed
  • Monitor trends: Weekly reviews prevent monthly surprises
  • Use alerts wisely: 50% warning, 80% action needed, 95% panic mode

​
🔒 Security First

  • Hash user IDs: Protect customer privacy
  • No secrets in tags: Attribution data is logged
  • Follow regulations: GDPR, CCPA compliance matters

​
📈 Analytics Optimization

  • Consistent naming: Same field names across all apps
  • Rich metadata: More context = better insights
  • Regular reviews: Monthly deep dives reveal optimization opportunities

​
Common Success Stories

🏢 Multi-Tenant SaaS
”We use App Attribution to bill customers accurately and identify our power users. Revenue attribution increased 40%.” 🚀 Product Teams
”Tracking by feature helped us kill underperforming products and double down on winners. ROI improved dramatically.” 🧑‍💻 Development Teams
”Separating staging from production costs was a game-changer. No more surprise bills from testing.” 📊 Analytics Teams
”Per-user tracking revealed usage patterns we never knew existed. Our retention strategies are now data-driven.” 💼 Finance Teams
”Cost allocation reports make budgeting and forecasting actually possible. CFO loves the transparency.”

​
What’s Next?

Ready to get surgical precision on your API costs? App Attribution gives you the visibility and control you need to optimize usage, allocate costs fairly, and make data-driven decisions about your AI investments. Start with a simple app_id on your next request, then gradually add more attribution fields as your needs grow. Your future self (and your finance team) will thank you. Happy tracking! 🚀
⌘I