Skip to content

Monitoring & Costs

Vortik gives you clear, actionable insights into how your AI models perform and how much they cost. Stop guessing and start making data-driven decisions.

Real-time Dashboard

The Vortik Dashboard is your command center. It gives you a live view of your app's behavior across all your environments and projects.

Key Metrics we track:

  • Volume: See requests per second, minute, or hour.
  • Latency (P50, P90, P99): Get detailed speeds for every single provider you use.
  • Error Rates: Catch API failures, rate limits, and timeouts the moment they happen.
  • Throughput: Visualize your token usage seamlessly.

Cost Management

Managing LLM spending across multiple providers can be a nightmare. Vortik makes it easy by bringing all your costs into one clear view.

Cost Tracking Features:

  • Aggregated Costs: See your total spend across all providers in real-time.
  • Model Breakdown: Quickly spot which models are burning your budget.
  • Project Allocation: Easily attribute costs to specific teams, environments, or projects.
  • Projections: Get an estimate of your future spending based on current patterns.

TIP

Vortik automatically calculates costs using the latest pricing from all major LLM providers. You don't have to configure any pricing sheets manually.

Observability and Troubleshooting

When something breaks, high-level metrics aren't enough. Vortik offers deep-dive tools for those tough debugging sessions.

Requests Log

Our Requests Log provides a searchable history of every single AI request passing through the gateway. Filter by project, model, provider, or status.

Every entry captures:

  • The full request and response payloads
  • Accurate token counts
  • Latency and exact costs
  • The associated OpenTelemetry trace (if you've enabled it)

Performance Benchmarks

Compare the speed and quality of different providers. Use these benchmarks to pick the perfect model for your latency, budget, or quality needs.


That's the core of the Vortik platform! If you have any questions or get stuck, feel free to reach out to our support team.