Pricing
One price per machine. Usage-based billing.
No itemized line items, no tier upsell, no discount waiting behind a sales call. Each machine type has one monthly price — compute, memory, disk, network and daily backups all included. The only add-on is additional storage, at a flat $1.00/GB.
Price your instance
Pick a machine and a disk. That's the price.
Slide through the size ladder and storage range. The estimate is computed live from the same cost model the console shows at checkout and the billing engine charges from.
Your machine type already includes 2 GiB of disk in its price. Anything on top is the only variable on your bill: a flat $1.00 per GB / month. Total disk caps at 4,096 GiB (4 TiB).
A replica set is N full instances — each one is a whole machine with its own copy of the data, so it bills N× the single price. Odd sizes only (a quorum of N tolerates ⌊N/2⌋ failures).
s1 instance 0.05 vCPU + 0.5 GB RAM + 2 GiB disk + network + daily backups | $9/mo |
Total what bills every month while running | $9/mo |
$9/mo is the whole price of a s1 instance — compute, memory, 2 GiB of disk, network and daily backups, nothing itemized separately. Additional storage is the only add-on, at $1.00/GB. Add members for HA and each one adds another $9/mo. Pause it and only the disk bills — $0.45/mo.
Tiers
Three ways to run it.
Free
For learning, prototypes and small graphs.
No card required
Start free- Cypher over Bolt 5.0–5.4
- Official Neo4j drivers
- MCP server
- Full-text search
- Import from object storage & warehouses
- Guaranteed IOPS allocation
- Daily backups with restore
- API keys & Terraform provider
- Audit log
- Dedicated node: unthrottled disk & CPU
- Organization SSO
- Custom terms & priority support
Shared
Most popularPacked on shared nodes with a guaranteed IOPS cap. The cheapest always-on graph database.
s1 through g4 · billed for what you use
Start free- Cypher over Bolt 5.0–5.4
- Official Neo4j drivers
- MCP server
- Full-text search
- Import from object storage & warehouses
- Guaranteed IOPS allocation
- Daily backups with restore
- API keys & Terraform provider
- Audit log
- Dedicated node: unthrottled disk & CPU
- Organization SSO
- Custom terms & priority support
Dedicated
Your own node: unthrottled disk and CPU, up to 512 GB RAM, daily backups.
g8 through g512 · billed for what you use
Start free- Cypher over Bolt 5.0–5.4
- Official Neo4j drivers
- MCP server
- Full-text search
- Import from object storage & warehouses
- Guaranteed IOPS allocation
- Daily backups with restore
- API keys & Terraform provider
- Audit log
- Dedicated node: unthrottled disk & CPU
- Organization SSO
- Custom terms & priority support
Enterprise
Custom terms, SSO, dedicated support and BYOC. Talk to us.
Custom sizing · talk to us
Talk to us- Cypher over Bolt 5.0–5.4
- Official Neo4j drivers
- MCP server
- Full-text search
- Import from object storage & warehouses
- Guaranteed IOPS allocation
- Daily backups with restore
- API keys & Terraform provider
- Audit log
- Dedicated node: unthrottled disk & CPU
- Organization SSO
- Custom terms & priority support
Rates
The whole invoice, in three lines.
A machine's bundled price, additional storage, and metered egress. Everything else is inside the machine price.
| Machine type | one bundled price, per month | from $9 |
| Additional storage | per GB / month, beyond the included disk | $1.00 |
| Internet egress | per GB, as used | $0.12 |
One price per machine type · additional storage $1.00/GB · internet egress $0.12/GB as used
Sizes
Every rung, with the price it computes to.
| Size | Tier | vCPU | Memory | Storage | Disk IOPS | Connections | Monthly |
|---|---|---|---|---|---|---|---|
| c0 | free | burst to 0.5 vCPU | 512 MB | 1 GB | up to 500 IOPS | 200 | $0 |
| s1 | shared | burst to 1 vCPU | 512 MB | 2 GB | up to 1,000 IOPS | 256 | $9/mo |
| s2 | shared | burst to 1 vCPU | 2 GB | 4 GB | up to 1,500 IOPS | 512 | $22/mo |
| s3 | shared | burst to 2 vCPU | 4 GB | 6 GB | up to 2,000 IOPS | 1,000 | $43/mo |
| g1 | shared | 1 vCPU | 1 GB | 8 GB | up to 2,000 IOPS | 512 | $68/mo |
| g2 | shared | 1 vCPU | 2 GB | 8 GB | up to 3,000 IOPS | 700 | $75/mo |
| g4 | shared | 1 vCPU | 4 GB | 8 GB | up to 4,000 IOPS | 1,000 | $90/mo |
| g8 | dedicated | 2 vCPU | 8 GB | 16 GB | Dedicated | 1,500 | $241/mo |
| g16 | dedicated | 3 vCPU | 16 GB | 32 GB | Dedicated | 2,000 | $295/mo |
| g24 | dedicated | 5 vCPU | 24 GB | 48 GB | Dedicated | 3,000 | $468/mo |
| g32 | dedicated | 6 vCPU | 32 GB | 64 GB | Dedicated | 4,000 | $587/mo |
| g48 | dedicated | 9 vCPU | 48 GB | 96 GB | Dedicated | 6,000 | $879/mo |
| g64 | dedicated | 12 vCPU | 64 GB | 128 GB | Dedicated | 8,000 | $1,170/mo |
| g128 | dedicated | 24 vCPU | 128 GB | 256 GB | Dedicated | 12,000 | $2,337/mo |
| g256 | dedicated | 48 vCPU | 256 GB | 512 GB | Dedicated | 16,000 | $4,671/mo |
| g384 | dedicated | 72 vCPU | 384 GB | 768 GB | Dedicated | 20,000 | $7,005/mo |
| g512 | dedicated | 96 vCPU | 512 GB | 1024 GB | Dedicated | 24,000 | $9,339/mo |
Questions
Pricing questions.
Is the price on this page what I actually pay?
Yes. The estimate here is produced by the same cost model the console uses at checkout and the same rates the billing engine charges. There is no separate list price and no negotiated discount you have to ask for.
What am I charged when an instance is paused?
Only its disk — the included disk plus any additional storage. Compute, memory, network and backups stop billing the moment the instance pauses, because billing follows actual usage, not the calendar month.
Is there a charge per query?
No. There is no per-query, per-operation or per-row fee. You pay for the resources the instance reserves, nothing else.
What about data transfer?
Traffic inside the region is included. Internet egress is billed as used at $0.12 per GB, so it appears on your invoice only if you pull data out of the region.
Can I change instance size later?
Yes. Sizes move up and down the ladder from g1 through g64 on Pro, and storage expands independently up to 4,096 GiB. Resizing does not require recreating the instance or changing your connection URI.
How do I get a Dedicated instance?
Dedicated covers g8 through g512, up to 96 vCPU and 512 GB of memory on isolated compute — your own node, with unthrottled disk and CPU. You can start one yourself from the console, or get in touch and we will size it with you.
Start now
~98.7%
token efficiency at 3,700 entities (see the footnotes above)
Start on the free tier.
One instance, no card, no expiry. Move up the ladder when the workload asks you to.
First-graph path
LiveCreate a free instance
No card. Ready in about a minute.
Connect your driver
bolt+s:// URI into the driver you already use.
Write two MERGEs
That's the entire shape of agent memory.
Point an agent at it
One MCP config block. No integration code.