Starter
$0/ then pay-as-you-go
Perfect for prototyping and small projects
- $100 free GPU credits
- A10G, A100 40GB, L40S GPUs
- Up to 10 deployments
- Community support
- 7-day log retention
Simple, transparent pricing
Pay per second of GPU usage. No hidden fees, no minimum commits. Scale up or down without penalty.
Choose your plan
All plans include autoscaling, multi-region deployment, and 99.95% uptime SLA.
Growth
$499/ /month + usage
For production teams shipping AI features
- Everything in Starter
- A100 80GB and H100 GPUs
- Unlimited deployments
- Priority support (4-hour SLA)
- 30-day log retention
- SSO and audit logs
- Private endpoints
Enterprise
Custom
For organizations with compliance and volume needs
- Everything in Growth
- Volume discounts (up to 35% off)
- Dedicated support engineer
- Custom SLA up to 99.99%
- HIPAA BAA and SOC 2
- VPC peering and BYOC
- Annual commit discounts
GPU compute pricing (per second)
- $0.00092
- A10G (24GB)
- $0.00278
- A100 40GB
- $0.00417
- A100 80GB
- $0.00833
- H100 (80GB)
Pricing questions
How does per-second billing work?
You're billed for the exact GPU time your model uses, rounded to the nearest second. If your inference takes 0.3 seconds, you pay for 1 second. Cold start time is not billed.
What happens when I hit my free credits?
Your deployments continue running and you're automatically switched to pay-as-you-go billing. You'll receive an email notification when you reach 80% of your credits.
Can I switch between GPU types?
Yes. You can change GPU types by updating your deployment config and redeploying. Pricing updates automatically.
Are there egress or data transfer fees?
No. All data transfer (ingress and egress) is included. You only pay for GPU compute time.
What payment methods do you accept?
We accept credit cards, ACH transfers, and wire payments for annual contracts. Enterprise customers can request invoicing.
Do you offer volume discounts?
Yes. Growth and Enterprise plans receive volume discounts starting at $5,000/month in usage. Contact our sales team for custom pricing.
Start with $100 in free GPU credits
No credit card required. Deploy your first model in minutes.