Billing & Refund Policy
In plain terms: RunbridgeAI runs on a prepaid credit model — you top up, we deduct based on actual usage. This page explains how credits work, what happens when a request fails, and how refunds are handled.
How Billing Works
RunbridgeAI uses a prepaid credit (Credits) system:
- Credits are available for use immediately after purchase and apply across all models on the platform.
- Usage is billed per request based on actual token consumption or generation units. There are no monthly fees or minimum spend requirements.
- Credits do not expire as long as your account remains in good standing.
- Enterprise customers may contact us for custom contracts and volume pricing.
Prices are listed on our Pricing page.
Auto Top-Up
You may enable automatic credit reloading in your account settings. When enabled, your selected payment method will be charged automatically when your credit balance falls below a threshold you configure.
- You are responsible for maintaining valid payment information on your account.
- RunbridgeAI is not liable for service interruptions caused by failed payments, bank declines, or disabled Auto Top-Up settings.
- Auto Top-Up can be modified or disabled at any time in your account settings.
Failed Generation Billing
Some requests may fail due to gateway rejection, upstream model errors, or timeouts. The following rules apply:
| Scenario | Charged? |
|---|---|
| Rejected at the gateway layer (invalid parameters, authentication failure, etc.) | No |
| Forwarded to upstream model; model returns an error (5xx) | No |
| Rejected due to policy-violating input (prompt content, reference images, etc.) | Reviewed on a case-by-case basis |
| Model begins generating but the request is interrupted mid-stream | Billed for tokens generated; exceptions may apply where the interruption is attributable to RunbridgeAI |
| Request completes successfully but output does not meet your expectations | Charged in full; requests related to significant platform-side quality degradation may be reviewed |
"Generation has begun" means the model has returned at least one token via streaming or a complete response.
Refund Policy
Purchased credits are generally non-refundable. If you have a special circumstance, please contact us at product@runbridge.ai — we review requests on a case-by-case basis and will work with you to find a reasonable resolution.
To help us process your request, please include: the amount topped up, the date of purchase, your current unused credit balance, and the reason for your request. We will respond within 5 business days.
RunbridgeAI reserves the right to approve or decline refund requests at its sole discretion. Approved refunds will be returned to the original payment method; non-refundable processing fees charged by third-party payment providers will be excluded from the refund amount.
Billing Disputes
If you believe you have been charged in error, submit a dispute within 30 days of the charge date via our Support page. Disputes raised after this period will be considered an acceptance of the charge to the fullest extent permitted by applicable law.
What RunbridgeAI Does Not Guarantee
- Credits are not personal property, are non-transferable, and have no cash equivalent value.
- Refunds are not a default entitlement and are subject to review and approval.
- The quality or suitability of generated outputs is outside the scope of this billing policy.
Support
For billing-related questions, visit: runbridge.ai/support
Last updated: June 2, 2026