Pricing
Our pricing model is resource based and designed to provide the capacity you need for every agent to run its own isolated Postgres branch without multiplying storage and compute costs. We don't bill you per user or per branch.
Open Source
Self-managed.
Built by the Xata team.
forever
Branching
IncludedScale-to-zero
IncludedWake-up time
~10sStorage
OpenEBS/MayastorAnonymization
StandardProvisioning
Self-managedSupport
CommunityObservability
Not includedCompliance
Not includedMulti-region
Not includedXata Cloud
Fully managed.
Pay as you go.
+ $0.28/GB/mo storage
Branching
IncludedScale-to-zero
IncludedWake-up time
~1sStorage
XatastorAnonymization
StandardProvisioning
Fully managedSupport
StandardObservability
IncludedCompliance
SOC 2, GDPRMulti-region
AWS, GCPBYOC
Your cloud, full control.
We manage the platform.
Get in touch with us if you are interested in learning more about BYOC deployments.
Branching
IncludedScale-to-zero
IncludedWake-up time
~1sStorage
XatastorAnonymization
Advanced (HIPAA)Provisioning
Your cloudSupport
24/7 + dedicated SEObservability
IncludedCompliance
SOC 2, GDPR + BAAMulti-region
AWS, Azure, GCPGive your agent our pricing context
Feed pricing.md to your agent for a complete, machine-readable pricing reference.
Pricing by region
Instance pricing
| Instance Type | vCPUs | RAM | Cost per Hour | Cost per Month |
|---|---|---|---|---|
| micro | ≤2 vCPU | 1 GB RAM | $0.012/hr | $9/month |
| small | ≤2 vCPU | 2 GB RAM | $0.024/hr | $18/month |
| medium | ≤2 vCPU | 4 GB RAM | $0.048/hr | $35/month |
| large | 2 vCPU | 8 GB RAM | $0.096/hr | $70/month |
| xlarge | 4 vCPU | 16 GB RAM | $0.192/hr | $140/month |
| 2xlarge | 8 vCPU | 32 GB RAM | $0.384/hr | $280/month |
| 4xlarge | 16 vCPU | 64 GB RAM | $0.768/hr | $561/month |
| 8xlarge | 32 vCPU | 128 GB RAM | $1.536/hr | $1121/month |
Storage
Companies and platforms, at any scale
Deploy via BYOC in your own cloud account, also available on the AWS Marketplace.
| Tier | Concurrent branches | Ideal for |
|---|---|---|
| Team | Up to 1,000 | Mature engineering orgs transitioning away from shared staging as agents become core to daily work |
| Company | Up to 10,000 | Centralized company-wide AI rollouts with agent governance policies |
| Platform | Up to 100,000 | Early stage platforms serving thousands of users with dedicated databases |
| Hyperscale | 100,000+ | Mature platforms provisioning databases for millions of users |
Frequently Asked Questions
Find answers to common questions about Xata pricing
Instance pricing (all regions)
us-east-1
| Instance | vCPUs | RAM (GiB) | USD/hour | USD/month (730h) |
|---|---|---|---|---|
| micro | <=2 | 1 | 0.012 | 9 |
| small | <=2 | 2 | 0.024 | 18 |
| medium | <=2 | 4 | 0.048 | 35 |
| large | <=2 | 8 | 0.096 | 70 |
| xlarge | 4 | 16 | 0.192 | 140 |
| 2xlarge | 8 | 32 | 0.384 | 280 |
| 4xlarge | 16 | 64 | 0.768 | 561 |
| 8xlarge | 32 | 128 | 1.536 | 1121 |
us-west-2
| Instance | vCPUs | RAM (GiB) | USD/hour | USD/month (730h) |
|---|---|---|---|---|
| micro | <=2 | 1 | 0.012 | 9 |
| small | <=2 | 2 | 0.024 | 18 |
| medium | <=2 | 4 | 0.048 | 35 |
| large | <=2 | 8 | 0.096 | 70 |
| xlarge | 4 | 16 | 0.192 | 140 |
| 2xlarge | 8 | 32 | 0.384 | 280 |
| 4xlarge | 16 | 64 | 0.768 | 561 |
| 8xlarge | 32 | 128 | 1.536 | 1121 |
eu-central-1 (1.15x base rate)
| Instance | vCPUs | RAM (GiB) | USD/hour | USD/month (730h) |
|---|---|---|---|---|
| micro | <=2 | 1 | 0.014 | 10 |
| small | <=2 | 2 | 0.028 | 20 |
| medium | <=2 | 4 | 0.055 | 40 |
| large | <=2 | 8 | 0.110 | 80 |
| xlarge | 4 | 16 | 0.221 | 161 |
| 2xlarge | 8 | 32 | 0.442 | 323 |
| 4xlarge | 16 | 64 | 0.883 | 645 |
| 8xlarge | 32 | 128 | 1.766 | 1289 |
Storage
USD 0.28 per GB per month. Powered by Xatastor NVMe. Single rate across all regions; regional compute multipliers do not apply to storage.
Cost model
Development (always included)
- Copy-on-write branching: branches share storage with the source database; only changed blocks are duplicated.
- Scale-to-zero: idle branches suspend compute entirely (USD 0/hr when idle) and can stay hibernated indefinitely.
- Production clone: an always-on instance that serves as the branch source, billed at the full instance rate times 730 hours/month.
Production on Xata (optional)
- Primary instance plus read replicas, all running 24/7.
- Storage replicated across all instances.
Branch storage delta model
Agent branches are short-lived and touch little data. The percentage of changed data per branch decreases as the production database grows.
| Database size | Estimated change data % | Delta per branch |
|---|---|---|
| 5 GB | 4.5% | 223 MB |
| 50 GB | 1.4% | 705 MB |
| 100 GB | 1.0% | 1.0 GB |
| 500 GB | 0.45% | 2.3 GB |
| 1 TB | 0.32% | 3.2 GB |
| 2 TB | 0.22% | 4.4 GB |
| 5 TB | 0.14% | 7.0 GB |
Formula: estimated_change_pct = 10 × database_size_gb^(-0.5), clamped to [0.05%, 10%].
Formulas
- Branch compute = instance_hourly_rate × active_hours_per_branch × branch_count
- Production clone compute = instance_hourly_rate × 730
- Production clone storage = database_size_gb × storage_rate_per_gb
- Branch delta storage = database_size_gb × storage_rate_per_gb × change_data_pct × branch_count
- Production compute (optional) = instance_hourly_rate × 730 × (1 + replica_count)
- Production storage (optional) = database_size_gb × storage_rate_per_gb × (1 + replica_count)
Constants and credits
- Hours per month: 730
- Typical active hours per branch: 1–20 hrs/month
- Onboarding credit: up to USD 100
- No per-branch, per-developer, or per-database fees.
Let's scope it for your team.
See how Xata fits your stack, your scale, and your security requirements.