Pricing
You pay for what you use — by the second for machines, by the byte for storage and bandwidth, by the token for AI. Where the leading cloud sells the same thing, its own published price is beside ours.
Everything that can be measured is metered and billed monthly in arrears for exactly what was used. No minimums, no commitments.
Support is people on a rota, available whether or not you call on them — so it is the one thing billed by the month.
A domain registered for a year, or prepaid credit to draw down. Paid once, when it is bought.
GPU instances
Per hour, billed by the second.
| Machine | Price | The leading cloud |
|---|---|---|
RTX 3090 24 GB VRAM · 8 vCPU · 24 GB RAM · us-east-1 | $0.5634/hr | $0.8048/hr30% less Its nearest GPU (L4) — it does not sell this one |
RTX 4090 24 GB VRAM · 8 vCPU · 32 GB RAM · us-east-1 | $0.7646/hr | $0.8048/hr5% less Its nearest GPU (L4) — it does not sell this one |
L40S 48 GB VRAM · 16 vCPU · 64 GB RAM · us-east-1 | $1.30/hr | $1.86/hr30% less Same GPU (L40S) |
A100 40GB 40 GB VRAM · 16 vCPU · 64 GB RAM · us-east-1 | $1.92/hr | $2.74/hr30% less Same GPU (A100 40 GB) |
A100 80GB 80 GB VRAM · 24 vCPU · 96 GB RAM · us-east-1 | $2.40/hr | $3.43/hr30% less Same GPU (A100 80 GB) |
A100 PCIe 80 GB VRAM · 16 vCPU · 96 GB RAM · us-east-1 | $2.40/hr | $3.43/hr30% less Same GPU (A100 80 GB) |
H100 SXM 80 GB VRAM · 24 vCPU · 96 GB RAM · us-east-1 | $4.82/hr | $6.88/hr30% less Same GPU (H100 80 GB) |
H100 PCIe 80 GB VRAM · 16 vCPU · 96 GB RAM · us-east-1 | $4.82/hr | $6.88/hr30% less Same GPU (H100 80 GB) |
CPU instances
Per hour, billed by the second. Compared with the leading cloud's cheapest machine that has at least as many vCPUs and as much memory.
CosmoBase
Database Each running database, by the hour. Two are included free. | $0.0112 / database-hour | $0.01630% less |
Database storage Data stored across your databases, metered by the GB-hour. 40 GB included free. | $0.1625 / GB-month | — |
Cloud Services
Object storage Objects stored, metered by the GB-hour. 10 GB included free. | $0.0218 / GB-month | $0.0235% less |
Panel
Build minutes Time spent building a deployment, by the second. | $0.007 / build minute | $0.0130% less |
Bandwidth Data served from your deployments. | $0.312 / GB | — |
CodeGraph
Repository storage Repository data, metered by the GB-hour. 100 GB included free. | $0.0273 / GB-month | — |
AI inference
General & code model — input Prompt tokens sent to the general, code and security models. | $0.105 / million tokens | $0.1530% less |
General & code model — output Tokens generated by the general, code and security models. | $0.455 / million tokens | $0.6530% less |
Included at no charge
On every account, every month. Usage beyond it is billed at the rates above.
- 2 databases
- 40 GB storage
- 10 GB object storage
- 10 repositories
- 100 GB repository storage
Support
Support is people on a rota, available whether or not you call on them. Billed monthly; cancel any time.
- · Email support, business hours
- · Response within 12 business hours
- · Architecture guidance
- · 24/7 email and chat
- · Production down: response within 1 hour
- · Unlimited contacts
- · Named technical account engineer
- · Business-critical: response within 30 minutes
- · Quarterly architecture and cost reviews
One-time
Prepaid credit is applied to your invoices automatically until it is used up. Domain registrations are quoted live per name, for the term you choose.
Applied automatically to your next invoices until it is used up. Does not expire.
Applied automatically to your next invoices until it is used up. Does not expire.
Applied automatically to your next invoices until it is used up. Does not expire.
Applied automatically to your next invoices until it is used up. Does not expire.
All prices in USD, before tax. Comparisons are with the leading cloud's own published on-demand rates for its US East region, as published 10 November 2026; where it tiers a price by volume, the comparison is with its first tier. See the price comparison for every figure.