Monitoring & Analytics

Comprehensive monitoring and analytics tools to track your DeepSeek API usage, performance, and costs in real-time.

Overview

DeepSeek provides powerful monitoring capabilities to help you:

Track API Usage: Monitor requests, tokens, and response times
Analyze Performance: Identify bottlenecks and optimization opportunities
Control Costs: Monitor spending and set budget alerts
Ensure Reliability: Track uptime and error rates
Optimize Applications: Gain insights into usage patterns

Dashboard Overview

Real-time Metrics

Access your monitoring dashboard at https://console.deepseek.com/monitoring

Key Metrics Display

Request Volume: Real-time API request counts
Response Times: Average and percentile response latencies
Token Usage: Input and output token consumption
Error Rates: Success/failure ratios and error types
Cost Tracking: Real-time spending and projections

Time Range Selection

Real-time: Live data updates every 30 seconds
Last Hour: Detailed minute-by-minute breakdown
Last 24 Hours: Hourly aggregated data
Last 7 Days: Daily summaries and trends
Last 30 Days: Weekly and monthly patterns
Custom Range: Flexible date range selection

Usage Analytics

Request Metrics

Volume Tracking

json

{
  "total_requests": 15420,
  "successful_requests": 15180,
  "failed_requests": 240,
  "success_rate": 98.44,
  "requests_per_minute": 127.5,
  "peak_requests_per_minute": 245
}

Model Usage Distribution

json

{
  "model_usage": {
    "deepseek-chat": {
      "requests": 8500,
      "percentage": 55.1,
      "tokens": 2150000
    },
    "deepseek-coder": {
      "requests": 4200,
      "percentage": 27.2,
      "tokens": 980000
    },
    "deepseek-vision": {
      "requests": 2720,
      "percentage": 17.7,
      "tokens": 1200000
    }
  }
}

Token Analytics

Token Consumption

Input Tokens: Tokens sent in requests
Output Tokens: Tokens generated in responses
Total Tokens: Combined input and output usage
Average Tokens per Request: Efficiency metrics
Token Rate Trends: Usage patterns over time

Token Usage Breakdown

json

{
  "token_usage": {
    "input_tokens": 1250000,
    "output_tokens": 3080000,
    "total_tokens": 4330000,
    "average_input_per_request": 81.1,
    "average_output_per_request": 199.7,
    "average_total_per_request": 280.8
  }
}

Performance Metrics

Response Time Analysis

Average Response Time: Mean latency across all requests
P50 Response Time: Median response time
P95 Response Time: 95th percentile latency
P99 Response Time: 99th percentile latency
Maximum Response Time: Slowest response recorded

Performance Trends

json

{
  "performance": {
    "average_response_time_ms": 245,
    "p50_response_time_ms": 180,
    "p95_response_time_ms": 450,
    "p99_response_time_ms": 850,
    "max_response_time_ms": 1200,
    "timeout_rate": 0.02
  }
}

Error Monitoring

Error Classification

Error Types

Authentication Errors: Invalid API keys or permissions
Rate Limit Errors: Quota exceeded or rate limiting
Validation Errors: Invalid request parameters
Server Errors: Internal service issues
Timeout Errors: Request timeout exceeded

Error Rate Tracking

json

{
  "error_analysis": {
    "total_errors": 240,
    "error_rate": 1.56,
    "error_types": {
      "authentication": 45,
      "rate_limit": 120,
      "validation": 35,
      "server_error": 25,
      "timeout": 15
    }
  }
}

Error Details

Error Response Format

json

{
  "error_id": "err_abc123",
  "timestamp": "2025-01-15T10:30:00Z",
  "error_type": "rate_limit",
  "error_code": "rate_limit_exceeded",
  "message": "Rate limit exceeded. Please try again later.",
  "request_id": "req_xyz789",
  "model": "deepseek-chat",
  "user_id": "user_123"
}

Error Trend Analysis

Error Rate Over Time: Track error patterns
Error Distribution: Breakdown by error type
Recovery Time: Time to resolve issues
Impact Analysis: Affected users and requests

Cost Monitoring

Billing Analytics

Cost Breakdown

json

{
  "cost_analysis": {
    "total_cost": 127.45,
    "currency": "USD",
    "billing_period": "2025-01",
    "cost_by_model": {
      "deepseek-chat": 68.20,
      "deepseek-coder": 35.15,
      "deepseek-vision": 24.10
    },
    "cost_by_token_type": {
      "input_tokens": 42.30,
      "output_tokens": 85.15
    }
  }
}

Cost Trends

Daily Spending: Track daily cost patterns
Monthly Projections: Forecast monthly expenses
Cost per Request: Efficiency metrics
Budget Utilization: Progress against set budgets
Cost Optimization: Recommendations for savings

Budget Management

Budget Alerts

json

{
  "budget_settings": {
    "monthly_budget": 500.00,
    "current_usage": 127.45,
    "usage_percentage": 25.49,
    "alerts": [
      {
        "threshold": 50,
        "status": "not_triggered"
      },
      {
        "threshold": 80,
        "status": "not_triggered"
      },
      {
        "threshold": 100,
        "status": "not_triggered"
      }
    ]
  }
}

Real-time Monitoring

Live Dashboard

WebSocket Connection

javascript

const ws = new WebSocket('wss://api.deepseek.com/monitoring/live');

ws.onmessage = function(event) {
  const data = JSON.parse(event.data);
  
  switch(data.type) {
    case 'request_count':
      updateRequestCounter(data.count);
      break;
    case 'response_time':
      updateLatencyChart(data.latency);
      break;
    case 'error_alert':
      showErrorAlert(data.error);
      break;
    case 'cost_update':
      updateCostDisplay(data.cost);
      break;
  }
};

Real-time Metrics

Live Request Counter: Real-time request volume
Response Time Graph: Live latency visualization
Error Rate Monitor: Real-time error tracking
Cost Meter: Live spending updates
System Health: Service status indicators

Alerting System

Alert Configuration

json

{
  "alerts": [
    {
      "name": "High Error Rate",
      "condition": "error_rate > 5%",
      "duration": "5 minutes",
      "channels": ["email", "webhook"],
      "enabled": true
    },
    {
      "name": "High Latency",
      "condition": "p95_response_time > 1000ms",
      "duration": "3 minutes",
      "channels": ["email", "slack"],
      "enabled": true
    },
    {
      "name": "Budget Alert",
      "condition": "monthly_cost > 80% of budget",
      "channels": ["email"],
      "enabled": true
    }
  ]
}

API Monitoring

Programmatic Access

Monitoring API Endpoints

bash

# Get usage statistics
curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.deepseek.com/monitoring/usage?period=24h"

# Get performance metrics
curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.deepseek.com/monitoring/performance?period=7d"

# Get cost analysis
curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.deepseek.com/monitoring/costs?period=30d"

Python SDK Integration

python

from deepseek import DeepSeek

client = DeepSeek(api_key="your-api-key")

# Get usage metrics
usage = client.monitoring.usage(period="24h")
print(f"Total requests: {usage.total_requests}")
print(f"Total tokens: {usage.total_tokens}")

# Get performance data
performance = client.monitoring.performance(period="7d")
print(f"Average response time: {performance.avg_response_time}ms")

# Get cost information
costs = client.monitoring.costs(period="30d")
print(f"Total cost: ${costs.total_cost}")

Custom Metrics

Custom Event Tracking

python

# Track custom events
client.monitoring.track_event(
    event_name="user_signup",
    properties={
        "user_id": "user_123",
        "plan": "pro",
        "source": "api"
    }
)

# Track custom metrics
client.monitoring.track_metric(
    metric_name="response_quality",
    value=4.5,
    tags={
        "model": "deepseek-chat",
        "user_type": "premium"
    }
)

Integration Examples

Grafana Dashboard

Grafana Configuration

json

{
  "dashboard": {
    "title": "DeepSeek API Monitoring",
    "panels": [
      {
        "title": "Request Volume",
        "type": "graph",
        "targets": [
          {
            "expr": "deepseek_requests_total",
            "legendFormat": "Total Requests"
          }
        ]
      },
      {
        "title": "Response Time",
        "type": "graph",
        "targets": [
          {
            "expr": "deepseek_response_time_p95",
            "legendFormat": "95th Percentile"
          }
        ]
      }
    ]
  }
}

Datadog Integration

Datadog Metrics

python

from datadog import initialize, statsd

# Initialize Datadog
initialize(api_key='your-datadog-api-key')

# Send custom metrics
statsd.increment('deepseek.requests.total')
statsd.histogram('deepseek.response_time', response_time)
statsd.gauge('deepseek.tokens.used', token_count)

Prometheus Metrics

Metrics Export

python

from prometheus_client import Counter, Histogram, Gauge

# Define metrics
REQUEST_COUNT = Counter('deepseek_requests_total', 'Total requests')
RESPONSE_TIME = Histogram('deepseek_response_time_seconds', 'Response time')
TOKEN_USAGE = Gauge('deepseek_tokens_used', 'Tokens used')

# Update metrics
REQUEST_COUNT.inc()
RESPONSE_TIME.observe(response_time)
TOKEN_USAGE.set(token_count)

Best Practices

Monitoring Strategy

Set Up Alerts: Configure alerts for critical metrics
Monitor Trends: Track long-term usage patterns
Cost Optimization: Regular cost analysis and optimization
Performance Tuning: Use metrics to optimize application performance
Capacity Planning: Plan for future usage based on trends

Performance Optimization

Response Time Monitoring: Track and optimize latency
Error Rate Analysis: Identify and fix error patterns
Token Efficiency: Optimize token usage for cost savings
Rate Limit Management: Monitor and manage rate limits
Caching Strategy: Implement caching based on usage patterns

Security Monitoring

Authentication Monitoring: Track authentication failures
Usage Anomalies: Detect unusual usage patterns
Access Patterns: Monitor API access patterns
Rate Limit Violations: Track rate limit violations
Error Pattern Analysis: Identify potential security issues

Troubleshooting

Common Issues

High Error Rates

Check authentication credentials
Verify request parameters
Monitor rate limits
Review server status

Performance Issues

Analyze response time trends
Check for rate limiting
Review request patterns
Optimize request parameters

Cost Overruns

Review usage patterns
Optimize token usage
Implement caching
Set budget alerts

Support Resources

Monitoring Documentation: /en/monitoring
API Reference: /en/api-reference
Support Portal: https://support.deepseek.com
Community Forum: https://community.deepseek.com

Next Steps

Set Up Monitoring: Configure your monitoring dashboard
Create Alerts: Set up critical alerts for your use case
Analyze Patterns: Review your usage patterns and trends
Optimize Performance: Use insights to improve your application
Plan Capacity: Use data for future capacity planning

Start monitoring your DeepSeek API usage today to ensure optimal performance, cost efficiency, and reliability for your applications.

Monitoring & Analytics ​

Overview ​

Dashboard Overview ​

Real-time Metrics ​

Key Metrics Display ​

Time Range Selection ​

Usage Analytics ​

Request Metrics ​

Volume Tracking ​

Model Usage Distribution ​

Token Analytics ​

Token Consumption ​

Token Usage Breakdown ​

Performance Metrics ​

Response Time Analysis ​

Performance Trends ​

Error Monitoring ​

Error Classification ​

Error Types ​

Error Rate Tracking ​

Error Details ​

Error Response Format ​

Error Trend Analysis ​

Cost Monitoring ​

Billing Analytics ​

Cost Breakdown ​

Cost Trends ​

Budget Management ​

Budget Alerts ​

Real-time Monitoring ​

Live Dashboard ​

WebSocket Connection ​

Real-time Metrics ​

Alerting System ​

Alert Configuration ​

API Monitoring ​

Programmatic Access ​

Monitoring API Endpoints ​

Python SDK Integration ​

Custom Metrics ​

Custom Event Tracking ​

Integration Examples ​

Grafana Dashboard ​

Grafana Configuration ​

Datadog Integration ​

Datadog Metrics ​

Prometheus Metrics ​

Metrics Export ​

Best Practices ​

Monitoring Strategy ​

Performance Optimization ​

Security Monitoring ​

Troubleshooting ​

Common Issues ​

High Error Rates ​

Performance Issues ​

Cost Overruns ​

Support Resources ​

Next Steps ​

Monitoring & Analytics

Overview

Dashboard Overview

Real-time Metrics

Key Metrics Display

Time Range Selection

Usage Analytics

Request Metrics

Volume Tracking

Model Usage Distribution

Token Analytics

Token Consumption

Token Usage Breakdown

Performance Metrics

Response Time Analysis

Performance Trends

Error Monitoring

Error Classification

Error Types

Error Rate Tracking

Error Details

Error Response Format

Error Trend Analysis

Cost Monitoring

Billing Analytics

Cost Breakdown

Cost Trends

Budget Management

Budget Alerts

Real-time Monitoring

Live Dashboard

WebSocket Connection

Real-time Metrics

Alerting System

Alert Configuration

API Monitoring

Programmatic Access

Monitoring API Endpoints

Python SDK Integration

Custom Metrics

Custom Event Tracking

Integration Examples

Grafana Dashboard

Grafana Configuration

Datadog Integration

Datadog Metrics

Prometheus Metrics

Metrics Export

Best Practices

Monitoring Strategy

Performance Optimization

Security Monitoring

Troubleshooting

Common Issues

High Error Rates

Performance Issues

Cost Overruns

Support Resources

Next Steps