Start free. Scale predictably. All prices in USD.
Try ReflexDB at no cost
For solo developers and small projects
20% off first 3 months with LAUNCH20
For growing engineering teams
20% off first 3 months with LAUNCH20
For mid-market teams that need compliance and control
20% off first 3 months with LAUNCH20
For regulated industries and high-volume workloads
Compute overage ($16/unit) is opt-in and disabled by default. Enable "burst scaling" in your account settings to allow usage beyond your included units. Data transfer beyond your included budget is billed at $0.12/GB.
| Feature | Free | Pro | Teams | Business | Enterprise |
|---|---|---|---|---|---|
| Price | $0 | $99/mo | $199/mo | $499/mo | from $1,599/mo |
| Included compute units | 0 (1 nano trial) | 10 | 20 | 50 | 150 |
| Max RAM per instance | 0.5 GB | 4 GB | 8 GB | 30 GB | 120 GB |
| Max databases | 1 | 10 | 25 | 50 | 100 |
| Included data transfer | 5 GB | 20 GB | 50 GB | 150 GB | 500 GB |
| Team seats | 1 | 3 | 10 | 25 | Unlimited |
| Build credits/month | 30 | 250 | 500 | 1,000 | 3,000 |
| Max tables per database | 25 | 100 | 250 | 400 | 400+ |
| Sync mode | Poll + CDC | Poll + CDC | Poll + CDC | Poll + CDC | Poll + CDC |
| Min sync interval (poll) | 300 s | 60 s | 30 s | 15 s | 5 s |
| SLA | None | 99.5% | 99.9% | 99.9% | 99.95%+ |
| Support | Docs | Email, 48 h | Email + Slack, 24 h | Email + Slack, 12 h | Dedicated Slack, 4 h |
| Multi-region deployment | — | — | ✓ | ✓ | ✓ |
| Custom subdomain | — | — | ✓ | ✓ | ✓ |
| Audit log export | — | — | — | ✓ | ✓ |
| Overage rate | — | $16/unit | $16/unit | $16/unit | $16/unit |
| Data transfer overage | — | $0.12/GB | $0.12/GB | $0.12/GB | $0.12/GB |
| Checkout | — | Stripe self-serve | Stripe self-serve | Stripe self-serve | Sales / invoice |
| Billing cadence | — | Monthly | Monthly | Monthly | Annual contract |
| Auto-scaling | — | — | — | — | ✓ |
| Rate limiting | ✓ | — | — | — | — |
| Auto-pause on inactivity | ✓ 2 days | — | — | — | — |
One compute unit equals one gigabyte of memory running continuously for a full calendar month. All instances use memory-optimised ARM64 Fargate with a 1:8 CPU:memory ratio — ideal for ReflexDB's in-memory workload. Sizes follow a clean power-of-2 CU chain.
Enterprise plans include auto-scaling — horizontal replicas behind a load balancer that scale automatically based on CPU utilisation. Lower tiers run a single task per instance.
Example: A Pro customer running one xlarge database (8 CU) plus one medium database (2 CU) fits exactly at the 10-CU baseline. Running two large databases with auto-scaling (16 CU) triggers overage for the excess 6 CU at $16/CU. 4xlarge and above are available on Business and Enterprise plans.
| Instance size | vCPU | Memory | Compute units |
|---|---|---|---|
| nano | 0.25 | 0.5 GB | 0.5 |
| small | 0.25 | 1 GB | 1 |
| medium | 0.25 | 2 GB | 2 |
| large | 0.5 | 4 GB | 4 |
| xlarge | 1 | 8 GB | 8 |
| 2xlarge | 2 | 16 GB | 16 |
| 4xlarge | 4 | 30 GB | 32 |
| 8xlarge | 8 | 60 GB | 64 |
| 16xlarge | 16 | 120 GB | 128 |
All prices are ex-VAT. Local taxes (e.g. VAT, GST, sales tax) may apply at checkout depending on your jurisdiction.