Know what every
workflow costs.
See exactly what every workflow, every provider, and every request costs. Set budget caps that actually stop runaway spend. Two-minute setup, no code.
No credit card · 14-day trial · 2-min setup
Works with your existing stack
The problem
Your AI costs are a black box.
The meter is always running.
A workflow looped on GPT-4. No caps, no alerts, no kill switch. Just a bill.
OpenAI shows one number. Anthropic another. No breakdown by client or workflow.
One key leaks, everyone is exposed. One rate limit takes down all clients.
You're justifying AI spend on vibes. Costs buried across 4 dashboards, no answer for your CFO.
Setup
Get your URL.
Paste it. You're done.
Pick your platform, name your workspace, and we'll generate the exact URL to paste into your workflow.
https://your-workspace.tokensense.io/chat/completionsPaste into: HTTP Request node → URL field
Do I need to change my API key?
No. Your existing OpenAI / Anthropic key stays the same. TokenSense passes it through.
Will my workflows break?
No. TokenSense is a transparent proxy — same request in, same response out. Your automations won't notice.
What about custom models?
TokenSense supports any OpenAI-compatible endpoint. Custom models, fine-tunes, and Azure deployments all work.
How it works
Watch a live request.
Every checkpoint. In real time.
See exactly what happens when your workflow fires an AI call through TokenSense. Every tag, budget check, and route — visible in milliseconds.
Waiting for request...
Features
Everything you need to
manage AI spend at scale.
Per-Workflow Cost Tracking
See exactly what each n8n workflow, Make scenario, or Zapier zap costs — broken down by client, by team, by project.
Budget Caps That Actually Stop
Set hard spending limits per workflow or per client. When the limit hits, AI calls stop immediately. No more surprise bills.
Real-Time Alerts
Get notified before budgets run out — not after. Alerts fire when spend approaches your limits.
Isolated Keys Per Project
Use separate AI keys per project, team, or client. Every workspace stays isolated — keys, budgets, and logs never mix.
Works With All Major AI Providers
Use OpenAI, Anthropic, Google AI, Mistral, or any compatible provider — all through one dashboard.
One Workspace Per Project
Give every project, team, or client its own isolated workspace. Their keys, budgets, and usage stay fully separate.
Full Usage Logs
Every AI call logged with the model used, how much it cost, and which workflow triggered it. Searchable and exportable.
What it fixes
Built for teams that ship
AI-powered automations.
You can't see where your AI spend is going
Provider dashboards show one number. You need the breakdown.
- See cost per workflow, per project, per team — not one lumped bill
- Attribute every dollar to the workflow that spent it
- Drill into any hour, day, or workflow — find the 3% of runs driving 80% of cost
- Export reports your finance team can actually read
Your AI bill keeps catching you off guard
Soft alerts don't stop spend. Hard caps do.
- Set hard limits per workflow, project, or team — requests stop when you hit them
- Real-time alerts at 50%, 80%, 100% of your budget
- Kill any runaway workflow in one click
- Sleep through the night without a $400 OpenAI surprise
One key, one provider, zero flexibility
Route to the right model. Rotate keys without breaking anything.
- Use OpenAI, Anthropic, Google, Mistral — one endpoint, one dashboard
- Separate keys per workflow, project, or team — isolated blast radius
- Swap models without editing a single workflow
- When one provider has an outage, route around it in 30 seconds
Why TokenSense
Built for automation teams.
Not another developer tool.
No servers, no infrastructure
TokenSense is a managed service. Nothing to deploy, nothing to maintain. Sign up and start tracking.
Predictable pricing — not usage-based
Free to start, paid plans from $29/mo. Unlimited workflows on every tier. No surprise bills from the tool itself.
Made for n8n, Make, and Zapier teams
Purpose-built for automation teams — agencies, ops teams, and solo builders running real workflows. Not for ML engineers or data scientists.
Your keys stay yours
Bring your own AI provider keys. Each workspace gets its own isolated keys and budgets.
2-minute setup
Change one URL in your workflow. That's the entire setup. No libraries, no plugins, no code.
Budget caps that actually stop spending
When a limit is hit, AI calls stop immediately. Not a warning email you see 3 hours later.
Pricing
Simple pricing that scales with you.
Start free. Upgrade when you need more capacity, longer retention, or advanced controls.
- 10,000 requests/month
- 2 team members
- 3 API keys
- 7-day log retention
- Unlimited projects & workflows
- CSV export
- Cost projections
- Per-key budgets
- 50,000 requests/month
- 10 team members
- 25 API keys
- 90-day log retention
- Unlimited projects & workflows
- CSV export
- Cost projections
- Per-key budgets
- 1,000,000 requests/month
- 50 team members
- Unlimited API keys
- 365-day log retention
- Unlimited projects & workflows
- PDF reports & audit log
- RBAC & key rotation
- Per-key budgets
Need more? Contact [email protected] for custom Enterprise plans with unlimited requests, SSO, and SLA.
FAQ
Common questions
From the Blog
Insights for teams running
AI in production
Stop guessing what AI costs.
Start controlling it.
Set up in 2 minutes. See which workflows and projects are driving your AI bill — immediately.
No credit card required · 14-day free trial