Get started with TokenSense

In 5 minutes you'll have a working TokenSense setup that tracks every AI call — cost, tokens, latency, and which workflow triggered it. No code changes to your automations.

What you'll need

• A TokenSense account (free tier available)
• An AI provider account (OpenAI, Anthropic, Google, Mistral, xAI, or others)
• An automation tool (n8n, Make, Zapier, or any HTTP-capable tool/code)

Step 1: Create your TokenSense account

Sign up at app.tokensense.io/signup. The free Starter tier includes 10,000 requests/month — enough to test and run lightweight workflows.

Step 2: Add your provider keys

Go to Provider Keysin the dashboard and add your AI provider's API key. TokenSense supports OpenAI, Anthropic, Google Gemini, Mistral, xAI, and more. TokenSense forwards your requests to the provider using your own key — you keep full control.

Step 3: Get your TokenSense API key

Go to Keysin the dashboard and create a new API key. This is the key you'll use in your automation tool instead of calling your provider directly.

Step 4: Use it in your tool

Point your tool at https://api.tokensense.io (your TokenSense endpoint) with your TokenSense API key. All requests flow through TokenSense, then to your provider — with full cost tracking.

What happens next

Your first request will appear in the TokenSense dashboard within seconds — with cost, token count, latency, and the model used. From there, set up projects to group usage and budgets to control spend.