2 minTroubleshooting

Billing & Credits

DAIRX uses a prepaid credit model. You're billed for actual runtime. Most billing issues resolve with the steps below.

Deployments blocked — "Insufficient credits"

DAIRX blocks new deployments when your credit balance hits $0. Running instances continue but new ones are refused.

Go to Dashboard → Billing → Add Credits and top up
Credits are available immediately after payment
Set up low-balance alerts at Billing → Alerts to avoid surprise blocks

Unexpected charges

GPU instances charge continuously, even when idle. The most common surprise is a running instance you forgot to terminate.

Check Dashboard → Instances → All for still-running instances
Terminate any instances you no longer need (billing stops immediately)
Review Billing → Usage for per-instance cost breakdown
Set spend alerts to get notified when hourly burn exceeds a threshold

⚠️ Common trap

JupyterLab instances look harmless but charge full GPU rate even with no cells running. Always terminate when done.

Credits not appearing after payment

Credits typically appear within 60 seconds of payment confirmation
Refresh the billing page — cached balance may be stale
Check your email for payment confirmation from Stripe
If payment was declined, check your card at Dashboard → Billing → Payment Methods

If credits don't appear after 5 minutes with a confirmed payment, contact support with your payment reference ID.

Spot instance charged more than expected

Spot prices fluctuate. The rate shown at deploy time is the starting rate — it can change during the instance lifetime.

DAIRX charges at the provider's real-time rate, not a locked-in price
For predictable billing, use on-demand or set spend alerts
Check Billing → Usage → Instance Details for the rate timeline

Export usage data

Go to Dashboard → Billing → Usage → Export
CSV includes: instance ID, provider, GPU type, start/end time, duration, cost
For invoices, check Billing → Invoices
Back to Docs