Managed database pricing looks messy because providers count different things: some bill per instance-hour, some flat per month, and almost nobody's headline number covers storage, backups, and standby nodes. The short answer: a fully managed PostgreSQL or MySQL database starts at $15 per month at Vultr, DigitalOcean, and Linode, while AWS, Google Cloud, and Azure advertise lower entry points on paper that climb much faster once you add real specs. This guide breaks down how each pricing model works, what the real entry prices are, every Vultr managed database plan currently published, and the add-ons that quietly grow most bills.
Quick take for readers who just want a number
Vultr's managed PostgreSQL and MySQL plans start at $15/month: 1 vCPU, 1 GB RAM, 25 GB storage, flat monthly billing with storage included
The ladder scales up to 24 vCPU / 128 GB for heavy production work, and every tier includes automated backups, encryption, and maintenance
If you later need high availability or read replicas, budget for a second node billed at the same rate
👉 Check Vultr's current managed database pricing
Pricing models differ more between providers than prices do, and mixing them up is how people end up with surprise bills. Four models cover almost the entire market.
Fixed monthly plans (Vultr, DigitalOcean, Linode): you pick a node size, pay one number per month, and storage is part of the deal. It's the easiest model to budget around, and it's what the plan list below uses.
Hourly instance plus storage (AWS RDS, Scaleway): compute meters by the hour and storage shows up as a separate line. Amazon's own pricing example puts a db.m5.xlarge Single-AZ instance at $0.410/hour, which the page multiplies out to $299.30 over a 730-hour month, before storage. The fine print matters too: RDS Custom has no minimum fee, bills partial hours in one-second increments, and applies a 10-minute minimum charge after status changes like restarts.
Azure splits its SQL Database billing into DTU (a bundled unit) or vCore (compute and storage unbundled), so the same workload can price out very differently depending on which you pick. Throughput-based NoSQL services skip node pricing entirely: DynamoDB charges per read/write capacity unit, Cosmos DB per Request Unit, Firestore per document operation. Analytics warehouses bill per scan, with BigQuery's on-demand rate at $6.25 per TB.
Licensing is one more variable on commercial engines. On Amazon RDS Custom for SQL Server, you can pay license-included rates or bring your own licenses through Microsoft's License Mobility program, and the hourly math changes substantially between the two.
The practical floor for a genuinely managed relational database is $15/month, and three providers sit exactly on it.
Vultr: $15/month for 1 vCPU, 1 GB RAM
DigitalOcean: $15/month for a single-node cluster with 1 GiB RAM, which its docs position for development and testing
Linode: $15/month for 1 vCPU, 1 GB
The Big 3 advertise lower entry points: AWS's db.t4g.micro lands around $12/month, GCP's db-f1-micro around $9, and Azure's cheapest tier around $5. Those are dev-grade machines, though, and storage usually bills separately, so the gap inverts quickly once workloads get real. One 2026 pricing guide that compared identical 4 vCPU / 16 GB PostgreSQL setups found roughly $280/month on AWS RDS and $260/month on Cloud SQL, with budget providers landing between about $120 and $185 depending on included storage. The same guide puts the smaller providers 50-80% below Big 3 pricing for the same engine.
One genuinely free option deserves mention: Oracle Cloud's Always Free tier includes a managed database with 1 OCPU and 20 GB, enough for small projects at $0. DigitalOcean also notes that features like point-in-time recovery and read-only nodes vary by engine, so check per engine before assuming anything at these entry prices.
Vultr runs managed PostgreSQL and MySQL on one shared plan ladder, plus a Redis-compatible managed caching service. Everything bills as a flat monthly rate in USD with storage included per plan, so there's no separate storage line to watch, and automated backups, encryption, and maintenance are part of the service. Adding a standby or read-replica node later means running a second node, so budget accordingly if availability matters. The tiers below reflect Vultr's published monthly rates; the pricing page remains the source of truth if anything has shifted.
For steady 24/7 workloads, this flat model with bundled storage almost always beats hourly-plus-storage on total cost. Here's the full ladder.
$15/month – 1 vCPU, 1 GB RAM, 25 GB storage
Flat monthly rate, storage included, backups and maintenance handled for you
Fits development, staging, and small production apps
👉 See the $15 plan on Vultr's pricing page
$30/month – 1 vCPU, 2 GB RAM, 38 GB storage
Same management features, double the memory and extra storage headroom
Fits busier small apps and light production duty
👉 View the $30 plan's current price
$60/month – 2 vCPU, 4 GB RAM, 75 GB storage
Two vCPUs for real query concurrency
A sensible default for production apps with steady traffic
👉 Check the $60 plan on Vultr's pricing page
$95/month – 2 vCPU, 8 GB RAM, 150 GB storage
Memory doubles for larger working sets and heavier caching
Fits data-driven production workloads
👉 See the $95 tier's current price
Past $95, the steps get bigger rather than incremental, with each tier jumping to a materially larger node.
$185/month – 4 vCPU, 16 GB RAM, 300 GB storage
The classic production database shape, with 300 GB of storage included rather than metered
Worth comparing directly against providers that bill storage as a separate line
👉 View the $185 plan on Vultr's pricing page
$300/month – 8 vCPU, 32 GB RAM, 600 GB storage
Fits high-traffic apps and multi-tenant workloads
👉 Check the $300 tier's current price
$600/month – 12 vCPU, 48 GB RAM, 900 GB storage
Dedicated-class capacity for larger teams and datasets
👉 See the $600 plan's current price
$950/month – 16 vCPU, 64 GB RAM, 1,050 GB storage
Fits memory-heavy production databases
👉 View the $950 tier on Vultr's pricing page
$1,500/month – 20 vCPU, 96 GB RAM, 1,320 GB storage
Large dedicated deployments with substantial storage built in
👉 Check the $1,500 plan's current price
$1,950/month – 24 vCPU, 128 GB RAM, 1,650 GB storage
Top of the current lineup at 24 vCPU and 128 GB
👉 See the $1,950 tier's current price
From $15/month – 1 GB of memory
In-memory store for sessions, queues, and rate limiting
Pairs with a Postgres or MySQL cluster when caching moves off your web servers; higher-memory tiers sit on the same pricing page
👉 Check Vultr's managed caching plans
The price on the page buys you the node. Four things grow the bill after that.
Engine stacking comes first. Managed databases sell per cluster, and every engine is its own cluster with its own minimum. DigitalOcean is the clearest documented case: its entry prices for Postgres, MySQL, and its Redis-compatible cache each land at about $15/month, so an ordinary three-engine app stack starts around $45/month before any real traffic. Vultr prices per cluster the same way, so count engines before counting savings.
Storage billed separately comes next. Scaleway charges €0.0993 per GB per month for its 5K block storage and €0.1489 for the 15K class. DigitalOcean adds $0.215 per GiB per month for extra storage on Standard clusters and $0.115 on Advanced, meaning 100 GB of extra Standard storage is $21.50/month on top of the node. This is the main context for Vultr's mid-tier pricing: the storage is inside the plan, so the monthly number is the whole node.
Backups are a third line item on some platforms. Scaleway bills backups and snapshots at €0.03 per GB per month, before tax. On Vultr and the other flat-rate providers, backup handling is built into the service instead.
High availability is the biggest multiplier. HA means a standby node, and standby nodes bill like primaries. DigitalOcean's docs make it concrete: HA "begins" at $30/month for the primary plus a matching $30/month standby, which is $60/month in practice, with read-only replicas starting at another $15/month. Whatever provider you pick, assume real redundancy roughly doubles the node cost. One offsetting detail at DigitalOcean: traffic to and from managed databases doesn't count against your bandwidth allowance, and it's worth checking whether your provider treats database egress the same way.
👉 See Vultr's full managed database plan list and current prices
For steady, predictable workloads, the budget providers win on price, and Vultr's flat-rate model is the easiest of the three to budget around. At the entry level there's little to separate them: Vultr, DigitalOcean, and Linode all start at $15/month with a 1 vCPU / 1 GB node.
The mid-tier is where the models diverge. That 4 vCPU / 16 GB production Postgres runs about $280/month on AWS RDS and $260/month on Cloud SQL, with storage billed on top of the instance. Among budget providers it lands between roughly $120 and $185/month depending on the provider and how much storage the plan includes, and Vultr's $185 tier carries 300 GB inside that price.
Where the competition genuinely beats Vultr is engine breadth. DigitalOcean also sells managed MongoDB, OpenSearch, and Kafka, with Kafka starting at $148.80/month because three brokers is the minimum sensible deployment. Vultr's current managed catalog sticks to PostgreSQL, MySQL, and Redis-compatible caching, so if you need those other engines as managed services, that's a real limit. The Big 3 remain the only home for their proprietary platforms: DynamoDB, Cosmos DB, Firestore, Aurora, AlloyDB, Redshift, and BigQuery. Budget pricing doesn't apply there because nothing equivalent exists elsewhere.
Azure's DTU-versus-vCore choice is its own trap. DTU bundles compute and storage into one unit, vCore unbundles them, and picking the wrong one for your workload can swing the bill in either direction.
👉 Check where Vultr's plans fit your budget
Match the billing model to the workload. Steady 24/7 databases belong on flat monthly plans; hourly and on-demand pricing only win for spiky, short-lived use.
Count engines, not just nodes. Three engines at $15 minimums is $45/month, so consolidating onto an engine you already pay for is often the biggest discount available.
Ask where storage and backups bill before comparing headline prices. A cheaper node with €0.0993/GB storage and €0.03/GB backups can cost more than a pricier plan with both included.
Postpone high availability until downtime actually costs you money. Standbys bill like primaries, so HA doubles the node line.
On DynamoDB, provisioned capacity with auto-scaling runs 60-80% cheaper than on-demand mode for predictable throughput.
Use genuinely free tiers for tiny projects: Oracle Cloud's Always Free managed database, Firestore's free daily operation allowances (50K reads, 20K writes), or Scaleway's 30-day trial on its smallest MongoDB node.
Around $15/month at Vultr, DigitalOcean, and Linode for a 1 vCPU / 1 GB managed MySQL or PostgreSQL node. Azure ($5), GCP ($9), and AWS (~$12) advertise less, but those are dev-grade tiers where storage and growth cost extra. Oracle Cloud's Always Free tier is the only genuinely $0 managed option worth naming.
Roughly double the node price, because a standby is a full second node. DigitalOcean's own numbers make it concrete: a $30/month primary plus a $30/month standby, $60/month all-in.
No. Each Vultr managed database plan bundles its storage into the flat monthly price, and backup handling is part of the service. That's the contrast with hourly-model providers like Scaleway, where storage (from €0.0993/GB/month) and backups (€0.03/GB/month) appear as separate lines.
For one production database with modest traffic, $15-60/month at Vultr covers most cases, and doubling that gives you a realistic HA figure. Add a second engine's minimum only if something you depend on truly requires it.
Automated backups, encryption, patching, failover tooling, and maintenance windows, which is the ops work you'd otherwise schedule around yourself. At $15/month, that's less than an hour of a database administrator's time for all of it.
Run this checklist against your own situation:
One Postgres or MySQL database for an app you're building → Vultr's $15 plan is enough, and every tier above it is a straight upgrade when real usage says so
Steady production traffic with heavy storage → Vultr's $95-$185 tiers, where storage is included instead of metered
Need managed MongoDB, Kafka, or OpenSearch → DigitalOcean's lineup covers those, while Vultr's current catalog sticks to PostgreSQL, MySQL, and caching
Need DynamoDB, Cosmos DB, Firestore, Aurora, or BigQuery-class warehousing → the Big 3, no way around it
Truly zero budget → Oracle's Always Free tier or Firestore's free allowances, accepting their limits
Two last checks before ordering anything: confirm the current price on the provider's own pricing page, since every number in this guide comes from pages updated in 2026 and this market reprices often, and decide now whether HA is a month-one requirement or a later upgrade, because it changes which tier makes sense.
👉 Open Vultr's pricing page and start on the $15 plan