Pricing
Pay only for what you use, billed by the hour.
Pricing Philosophy
SailOps pricing is designed to be:
- Transparent - No hidden fees or surprises
- Fair - Pay only for actual usage
- Granular - Billed per hour
- Scalable - Costs scale with your business
Containers
Compute Classes
Containers are billed based on compute class and the resources you configure. Each compute class has a fixed CPU:RAM ratio, and you adjust resources using sliders within the class limits.
| Class | CPU:RAM Ratio | Max CPU | Max RAM | Description |
|---|---|---|---|---|
| Eco | 1:2 | 3 cores | 6 GB | Development and testing |
| Standard | 1:4 | 15 cores | 58 GB | General production workloads |
| Spot | 1:2 | 3 cores | 6 GB | Best-effort spot capacity, may be unavailable when spare capacity runs out |
| Production | 1:4 | 6 cores | 28 GB | Critical workloads with dedicated resources |
| Prod. Compute | 1:2 | 6 cores | 16 GB | CPU-intensive tasks (data processing, APIs) |
| Prod. Memory | 1:8 | 6 cores | 58 GB | Memory-intensive tasks (caching, ML inference) |
| ARM | 1:4 | 8 cores | 32 GB | Cost-optimized ARM-based workloads |
Within each class, configure resources using sliders:
- CPU: Adjust in increments (varies by class)
- Memory: Automatically calculated based on CPU:RAM ratio
Container Billing
Containers are billed:
- Per hour of runtime
- For each replica running
- Based on the compute class and resources selected
Example:
Container: api-server
Class: Standard
Resources: 2 vCPU, 8 GB RAM (1:4 ratio)
Replicas: 3
Uptime: 720 hours/month
Cost = (class rate × resource allocation) × 3 replicas × 720 hours
GPU Compute Classes
GPU compute classes are billed separately for GPU resources in addition to standard CPU/RAM billing. GPU billing is calculated per GPU-fraction-hour, with each class offering different GPU models and allocation modes.
| Class | GPU Model | GPU Mode | GPU Memory | Price Dimension | Description |
|---|---|---|---|---|---|
| gpu-l4 | NVIDIA L4 | Dedicated | 24 GB | Per GPU-hour | General-purpose GPU for AI inference |
| gpu-h100-mig-small | NVIDIA H100 | MIG 1g.10gb | 10 GB | Per slice-hour | H100 MIG slice, minimum billing duration applies |
| gpu-h100-mig-medium | NVIDIA H100 | MIG 2g.20gb | 20 GB | Per slice-hour | H100 MIG slice, minimum billing duration applies |
| gpu-l40s | NVIDIA L40S | Dedicated | 48 GB | Per GPU-hour | High-memory GPU for large models |
| gpu-h100 | NVIDIA H100 | Dedicated | 80 GB | Per GPU-hour | Top-tier GPU for training and inference |
GPU Billing Notes:
- GPU classes enforce minimum CPU and RAM requirements
- MIG classes have minimum billing durations to prevent abuse
- Some GPU classes require BUSINESS or CUSTOM support plans
- GPU utilization and memory metrics are tracked separately
Autoscaling Costs
When autoscaling adds replicas:
- Only active replicas are billed
- Scaling down stops billing immediately
- Min replicas always billed
Databases
PostgreSQL Pricing
PostgreSQL databases use flavors within compute classes. Each compute class (eco, standard, production) offers a set of predefined flavors (small, medium, large, xlarge) with fixed CPU and RAM allocations.
Select a compute class, then choose a flavor that matches your workload requirements. Storage is configured separately.
MariaDB Pricing
MariaDB databases follow the same model as PostgreSQL: select a compute class (eco, standard, production), then choose a flavor (small, medium, large, xlarge). Storage is configured independently.
MariaDB supports:
- Multiple databases per instance
- Public access for external connections
- Automated backups
Database Flavors
Both PostgreSQL and MariaDB use the following structure:
| Compute Class | Available Flavors | Description |
|---|---|---|
| Eco | small, medium, large, xlarge | Development and testing |
| Standard | small, medium, large, xlarge | General production databases |
| Production | small, medium, large, xlarge | Critical databases with dedicated resources |
Each flavor defines a fixed CPU and RAM allocation. Exact resource values for each flavor are available via the Pricing API.
Storage Classes
| Class | Max IOPS | Best For |
|---|---|---|
| standard | 5,000 | General workloads, development |
| performance | 15,000 | High I/O, OLTP, analytics |
All storage uses NVMe SSDs. Storage can be increased at any time without downtime.
High Availability
HA doubles compute costs (standby replica running synchronously):
- Primary instance: full compute cost
- Standby replica: additional compute cost
- Automatic failover in < 30 seconds
- Storage shared and replicated
NATS Messaging
NATS Cluster Pricing
NATS clusters use flavors within compute classes, similar to databases. Select a compute class (eco, standard, production), then choose a flavor (small, medium, large, xlarge).
| Compute Class | Available Flavors | Description |
|---|---|---|
| Eco | small, medium, large, xlarge | Development and testing |
| Standard | small, medium, large, xlarge | General production messaging |
| Production | small, medium, large, xlarge | Critical messaging with dedicated resources |
NATS clusters are billed per-hour based on the selected compute class, flavor, and number of replicas.
Cache
Valkey/Redis Instances
Cache instances use flavors within compute classes, following the same model as databases and NATS. Select a compute class (eco, standard, production), then choose a flavor (small, medium, large, xlarge).
| Compute Class | Available Flavors | Description |
|---|---|---|
| Eco | small, medium, large, xlarge | Development and testing |
| Standard | small, medium, large, xlarge | General production caches |
| Production | small, medium, large, xlarge | Critical caches with dedicated resources |
Cache instances are billed per-hour based on the selected compute class, flavor, and number of replicas.
Websites
Hosting Plans
Websites use fixed plans built on the Eco compute class:
| Plan | CPU | Memory | Replicas | Best for |
|---|---|---|---|---|
| Starter | 0.25 vCPU | 256 MiB | 1 | Personal sites, prototypes |
| Standard | 0.5 vCPU | 512 MiB | 2 | Business sites, moderate traffic |
| Pro | 1 vCPU | 2048 MiB | 3 | High-traffic sites, PHP applications |
Prices shown via the API are per instance; billing multiplies by the plan's replica count at Eco class rates.
Website Storage
Website files are stored in S3. Storage is billed per GiB per hour regardless of whether the website is running.
Start/Stop
Stopped websites are not charged for compute. S3 storage charges still apply.
Gateways
Gateway Classes
Gateways are billed per-hour based on the selected class:
| Class | Bandwidth | Best For |
|---|---|---|
| gw-small | 200 Mbps | Development, small applications |
| gw-medium | 500 Mbps | Medium traffic applications |
| gw-large | 1 Gbps | High-traffic production |
| gw-xlarge | 4 Gbps | Enterprise, very high traffic |
Included Features
| Feature | Included |
|---|---|
| Custom domains | Free (unlimited) |
| SSL certificates | Free (auto-provisioned) |
| Routes | Free (unlimited) |
Managed Observability
Managed observability usage covers telemetry sent through SailOps OpenTelemetry ingestion and retained in the managed observability backends. It is separate from normal resource compute billing.
Included telemetry
| Plan | Metric samples / month | Trace spans / month | Telemetry storage | Active series | Metric retention | Metric sample rate limit |
|---|---|---|---|---|---|---|
| Basic | 5 million | 1 million | 10 GiB-month | 500 | 7 days | 100 samples/s |
| Starter | 50 million | 5 million | 25 GiB-month | 10,000 | 30 days | 1,000 samples/s |
| Business | 500 million | 25 million | 100 GiB-month | 100,000 | 90 days | 10,000 samples/s |
| Custom | 0 included | 0 included | 0 included | Custom | 365 days | Custom |
Overage pricing
| Usage dimension | Price |
|---|---|
| Metric samples | €0.10 per million samples |
| Trace spans | €0.30 per million spans |
| Telemetry storage | €0.08 per GiB-month |
For Basic, metric ingestion is hard-capped when the included metric sample quota is reached. Paid plans continue ingesting and bill overage. Custom plans bill managed observability from the first unit unless your contract defines custom terms.
You can monitor current usage, quota, projected overage, active series, and per-project breakdown in Monitoring → Usage → Telemetry.
Volumes
All persistent volumes — block volumes attached to containers and shared (RWX) volumes — are billed per GiB per hour based on provisioned capacity.
Block volumes
Block volumes are the persistent volumes attached to a container. They are billed on the provisioned size of each volume, using the storage class rate:
| Storage Class | Description |
|---|---|
| standard | General-purpose NVMe (5K IOPS) |
| performance | High-throughput NVMe (15K IOPS) |
- Billing starts when the volume is first bound
- Billing stops when the volume is deleted
- Detached volumes continue to be billed until explicitly deleted
- Resizing a volume updates the billing rate immediately
Shared volumes
Shared (RWX) volumes are billed per GiB per hour based on the provisioned size, regardless of actual usage. Shared volumes persist independently of any container lifecycle.
- Billing starts when the volume is created
- Billing stops when the volume is deleted
- Resizing a volume updates the billing rate immediately
Object Storage
Object Storage is billed on metered usage across four dimensions. There is no fixed cost — you pay only for what you consume.
| Dimension | Price (excl. tax) | Description |
|---|---|---|
| Storage | €0.79 per GiB-month | Time-weighted average of stored bytes, including all object versions |
| Writes & listings | €5.00 per million requests | PUT, POST, DELETE and list operations |
| Reads & HEAD | €0.50 per million requests | GET and HEAD operations |
| Downloads | €0.02 per GiB | Egress bytes — same rate from the SailOps internal network and the Internet |
Quantities are billed fractionally: there is no minimum charge and requests are not rounded up to the next million.
Usage is measured per bucket and aggregated at the project level. Costs appear in the Object Storage tab of the billing dashboard.
Object Storage requires an activated billing account and is unavailable during the free trial.
Cost Estimation
Costs depend on your selected compute classes, resources, and usage. Here are example configurations:
Example: Development Setup
Container (Eco, 0.5 vCPU / 1 GB × 1 replica):
720 hours/month
Database (Eco, 0.5 vCPU / 1 GB):
720 hours/month + 10 GB storage
Cache (Eco, 0.5 vCPU / 1 GB):
720 hours/month
Gateway (gw-small):
720 hours/month
Example: Production Setup
Container (Standard, 2 vCPU / 8 GB × 3 replicas):
720 hours/month
PostgreSQL Database (Standard, medium flavor + HA):
720 hours/month + 50 GB performance storage
MariaDB Database (Eco, small flavor):
720 hours/month + 10 GB standard storage
NATS Cluster (Standard, small flavor × 3 replicas):
720 hours/month
Cache (Standard, small flavor):
720 hours/month
Gateway (gw-medium):
720 hours/month
View current pricing in the dashboard when creating resources.
Billing Cycle
Monthly Billing
- Billing period: 1st to last day of month
- Invoice generated on 1st of following month
- Payment due within 14 days
Real-Time Usage
View current usage anytime:
- Click Billing in the left sidebar
- Select a resource tab:
- Summary — Overall cost overview
- Containers — Container compute costs
- Websites — Website hosting costs
- Gateways — Gateway bandwidth costs
- PostgreSQL — PostgreSQL database costs
- MariaDB — MariaDB database costs
- Caches — Valkey/Redis cache costs
- NATS — NATS messaging costs
- Jobs — One-time job costs
- Cron Jobs — Scheduled job costs
- Marketplace — Costs of existing instances deployed before retirement
- Custom Services — External service costs
- Volumes — Block and shared volume storage costs
- Object Storage — Object storage usage costs
- See accumulated costs and usage details for the selected resource type
Payment Methods
Accepted Methods
- Credit/Debit Card — Visa, Mastercard, and other networks supported by Stripe
- SEPA Direct Debit — Bank transfers for European bank accounts
Payments are processed securely through Stripe. Card and bank details are never stored on SailOps servers.
Adding Payment Method
- Open Organization Settings (click your Organization name at bottom left > Settings)
- Select the Payments tab
- Click Add Payment Method
- Enter card details in the Stripe secure form
- The card is set as your default payment method
Stripe Customer Portal
You can manage your payment methods and billing details directly through the Stripe Customer Portal:
- Go to Organization Settings > Payments tab
- Click Manage Billing to open the Stripe portal
- From there you can update card details, view payment history, and download receipts
Cost Optimization
1. Right-Size Resources
Match compute class to actual needs:
If CPU < 30% and Memory < 50% consistently:
→ Consider downgrading compute class
2. Use Autoscaling
Scale down during low traffic:
autoscaling:
min: 1 # Low traffic minimum
max: 10 # Peak capacity
3. Clean Up Unused Resources
Regularly review and remove:
- Stopped containers
- Unused databases
- Old backups
4. Use Cache Effectively
Cache reduces database load:
- Smaller database needed
- Fewer queries = faster response
- Lower costs overall
5. Optimize Storage
- Compress data when possible
- Clean up old logs
- Use appropriate retention policies
FAQ
How am I charged for stopped containers?
Stopped containers:
- No compute charges
- Storage still charged if persistent
What happens if payment fails?
- Email notification sent immediately
- Grace period starts — your services continue running
- Stripe retries the payment automatically
- If the invoice remains unpaid after the grace period, your organization is suspended
- Update your payment method to restore services instantly
Can I set spending limits?
SailOps uses pay-as-you-go billing. To control costs:
- Stop or suspend unused resources and projects
- Monitor usage in the Billing dashboard
- Contact support for custom billing arrangements on the CUSTOM plan
Are there setup fees?
No. SailOps has no:
- Setup fees
- Minimum commitments
- Cancellation fees
How do I get an invoice?
Invoices are:
- Generated monthly
- Available in Billing (click in sidebar) > Summary tab
- Sent to billing email
Pricing API
SailOps exposes a pricing endpoint that returns current rates for all resource types, scoped to your organization.
GET /api/v1/organizations/{organizationId}/pricing
If your organization has a CUSTOM plan with negotiated rates, the pricing endpoint returns your contract-specific prices. The billing dashboard and cost estimators use these same rates, so the amounts you see always match your invoice.
The response includes:
| Field | Description |
|---|---|
computeClasses | Container compute class pricing (per-CPU-hour, per-RAM-GiB-hour) |
storageClasses | Storage pricing (per-GiB-hour, max IOPS) |
loadBalancerClasses | Gateway pricing (per-hour, bandwidth) |
databaseClasses | Database and service flavor pricing (per-hour per flavor) |
websitePlans | Website plan pricing (per-hour per instance, CPU, memory) |
Additional fields for GPU compute classes:
| Field | Description |
|---|---|
minCpu | Minimum CPU cores required for the class |
minRamMiB | Minimum RAM in MiB required for the class |
gpuType | GPU model (L4, L40S, H100) |
gpuMode | GPU allocation mode (dedicated, mig-1g.10gb, mig-2g.20gb) |
gpuMemoryGiB | GPU memory per unit in GiB |
pricePerGpuHour | Price per GPU-fraction-hour |
maxGpus | Maximum number of GPUs per container |
minBillSeconds | Minimum billing duration for MIG classes (anti-abuse) |
requiresPlans | Array of required support plans (BUSINESS, CUSTOM) |
Use this endpoint to build cost estimators or integrate pricing into your CI/CD pipelines. The full schema is defined in the Public API documentation.