Cheapest managed database for a side project

A small production database that is idle most of the time.

List prices checked

How this is scored

Ranked on entry price, but the split that matters is architectural: serverless engines that scale to zero cost nothing while idle, while provisioned instances bill every hour regardless. For a side project that difference usually beats the sticker price.

All 11, ranked for this workload

#ServiceProviderCost for this workloadNotes
1TursoTurso$8.25/mo entry · has free tier500 databases, 9 GB of storage, and 1 billion row reads per month
2MongoDB AtlasMongoDB$9.00/mo entry · has free tierM0 shared cluster with 512 MB, always free
3Google Cloud SQLGoogle$9.37/mo entry
4Upstash RedisUpstash$10/mo entry · has free tier500,000 commands per month
5AWS RDS for PostgreSQLAmazon Web Services$12/mo entry · has free tier750 hours per month of a db.t-class instance for 12 months
6Azure Database for PostgreSQLMicrosoft$12/mo entry · has free tier750 hours per month of a burstable instance for 12 months
7Scaleway Managed DatabaseScaleway$13/mo entry
8Neon Serverless PostgresNeon$19/mo entry · has free tier0.5 GB of storage and 190 compute hours per month
9SupabaseSupabase$25/mo entry · has free tier500 MB database and 5 GB of egress per month
10CockroachDB CloudCockroach Labs$25/mo entry · has free tier$0 basic tier with monthly request-unit and storage allowances
11PlanetScalePlanetScale$39/mo entry · has free tierNone; the Hobby tier was retired in April 2024

Published list prices only. Rows that do not publish a rate for an axis this workload depends on are excluded rather than shown as cheap.

Questions

Cheapest managed database for a side project in 2026?
Turso from Turso — $8.25/mo entry · has free tier. MongoDB Atlas is second at $9.00/mo entry · has free tier. This ranking is specific to the workload described above; a different usage shape reorders it.
How is this ranking weighted?
Ranked on entry price, but the split that matters is architectural: serverless engines that scale to zero cost nothing while idle, while provisioned instances bill every hour regardless. For a side project that difference usually beats the sticker price.
Why does this differ from the general cheapest list?
Because a single price axis never describes a real workload. Ranking by headline rate answers "who is cheapest per unit"; this page answers "who is cheapest for this job", and the two orders are often very different — which is the whole reason it exists as a separate table.
What is not accounted for?
Committed-use discounts, regional price variation, per-request charges, support plans, and minimum retention. These are published list prices applied to one stated workload — a shortlist to price properly with the vendor, not a quotation.

Other workloads