Vercel AI Gateway usage and spend
Vercel AI Gateway observability shows request volume, latency, token usage, and spend across providers in one dashboard.
Hub: Tokenminning with Vercel AI Gateway.
Open the dashboard
- Go to vercel.com/dashboard
- Select your project → AI Gateway → Observability
- Filter by model, provider, and time range
See observability docs for metrics reference.
Usage and billing
Usage and billing explains pricing: zero markup on tokens. BYOK uses your provider keys at provider rates.
| Mode | Cost |
|---|---|
| AI Gateway key | Provider rates, no markup |
| BYOK | Your provider contract rates |
App attribution
Track which apps drive spend with app attribution headers. Tag requests by feature, route, or deployment.
Separate environments
| Environment | Pattern |
|---|---|
| Preview | Separate AI Gateway key or budget alert |
| Production | Production key with spend monitoring |
| Local dev | Direct provider keys or local Ollama |
Spending guardrails
- Review observability weekly — which model slug dominates?
- Set alerts before monthly budget thresholds
- Use BYOK only when enterprise contracts beat gateway pooling
- Attribute spend per app with attribution headers
Related
Last updated on