Your first database outlives most other early technical choices, because data is hard to move. The good news is that managed options are cheap at small scale: in AWS Mumbai, RDS PostgreSQL and MySQL on db.t4g.micro both cost $0.021 an hour, and DynamoDB includes 25 GB of free storage. This guide compares four common choices with prices checked on 7 November 2026.
Start with your data, not the brand
Most business software, such as CRMs, billing, booking and healthcare records, is relational: customers have orders, orders have line items, and you need reports that join them. A relational database such as PostgreSQL or MySQL handles this well and enforces correctness with transactions and constraints.
Document and key-value databases such as MongoDB and DynamoDB shine when access patterns are known in advance, data is naturally nested, or you need very high scale with predictable latency. The wrong choice usually shows up later as complex application code that recreates joins or transactions by hand.
PostgreSQL
PostgreSQL is an open-source relational database with strong standards support, JSON columns, full-text search and a large extension ecosystem, including vector search for AI features. It is offered as a managed service by every major cloud, so you are not locked in.
In AWS Mumbai, RDS PostgreSQL costs $0.021 an hour on db.t4g.micro, $0.042 on db.t4g.small and $0.084 on db.t4g.medium in a single zone, with gp3 storage at $0.131 per GB-month. On Azure Central India, PostgreSQL Flexible Server B1ms is $0.0245 an hour and B2s is $0.098.
MySQL
MySQL is the other widely used open-source relational database and powers many web frameworks and content systems. It is simple to run, well understood by hosting providers, and fast for read-heavy web workloads. Large pools of developers know it, and most backup, migration and monitoring tools support it out of the box, which lowers the cost of hiring and operations for a small company.
Managed pricing is the same as PostgreSQL on AWS: db.t4g.micro MySQL in Mumbai is $0.021 an hour in one zone and $0.042 with Multi-AZ. For a new product with no existing MySQL code, PostgreSQL usually offers more features for the same price; if your framework or team already uses MySQL, it is a perfectly sound choice. Aurora, AWS's cloud-built engine, supports both, with Aurora PostgreSQL Serverless v2 billed at $0.18 per ACU-hour in Mumbai.
MongoDB and compatible services
MongoDB stores flexible JSON-like documents, which makes early iteration quick when the data shape changes often. It suits content catalogues, event data and products where each record is read and written as a whole. It is less suited to data with many relationships that must stay consistent, such as invoices and payments.
Besides MongoDB's own Atlas service, the big clouds offer compatible managed options. Amazon DocumentDB in Mumbai costs $0.10961 an hour on db.t4g.medium (about $80.02 a month) plus $0.11 per GB-month of storage. Azure Cosmos DB serverless, which offers a MongoDB-compatible API, charges $0.285 per million request units in Central India.
DynamoDB
DynamoDB is AWS's serverless key-value and document database. There are no servers to size: in on-demand mode you pay per request, with Mumbai prices of $0.71 per million writes and $0.1425 per million reads, plus $0.285 per GB-month after the first 25 GB.
As an example, 10 million writes and 50 million reads a month with 10 GB stored costs about $14.23. The catch is design: you must plan access patterns and keys up front, ad hoc reporting is hard, and moving off DynamoDB later means rewriting the data layer.
Price comparison for a small product
The table compares entry options at 730 hours a month, using the AWS Price List for Mumbai and the Azure Retail Prices API for Central India on 7 November 2026. Storage, backups beyond free allowances and data transfer are extra.
| Option | Type | Unit price | Approx. monthly |
|---|---|---|---|
| RDS PostgreSQL db.t4g.micro | Relational | $0.021/hour | $15.33 |
| RDS MySQL db.t4g.micro | Relational | $0.021/hour | $15.33 |
| RDS PostgreSQL db.t4g.medium Multi-AZ | Relational, HA | $0.167/hour | $121.91 |
| Aurora PostgreSQL Serverless v2 | Relational, serverless | $0.18 per ACU-hour | Depends on load |
| Azure PostgreSQL Flexible B1ms | Relational | $0.0245/hour | $17.89 |
| Amazon DocumentDB db.t4g.medium | Document | $0.10961/hour | $80.02 |
| DynamoDB on-demand | Key-value | $0.71/M writes, $0.1425/M reads | Usage based |
| Azure Cosmos DB serverless | Multi-model | $0.285 per million RUs | Usage based |
A practical recommendation
For most first products, choose managed PostgreSQL. It handles relational data, flexible JSON fields and search in one place, runs on every cloud, and costs about $15 to $18 a month at the smallest size in India. Add a cache or a specialised store only when a measured bottleneck appears. Keep database exports in S3 Standard, which costs $0.025 per GB-month in Mumbai, and use credits such as AWS Activate Founders ($1,000 to start) to cover year one.
If you need zone-redundant databases on Azure, pick Central India (Pune) or India South Central (Hyderabad), the two Indian Azure regions with three availability zones. Choose DynamoDB when you are committed to AWS, your access patterns are simple and known, and you want no database servers at all. XDQ Labs defaults to PostgreSQL for new SaaS builds and adds DynamoDB or a document store for specific high-volume features.
- Relational business data: PostgreSQL (or MySQL if your stack already uses it)
- Simple, very high-volume key lookups on AWS: DynamoDB
- Rapidly changing document data: MongoDB or a compatible service
- Always: automated backups and a tested restore
Frequently asked questions
Is PostgreSQL better than MySQL for a startup?
Both are reliable and cost the same on AWS RDS in Mumbai, $0.021 an hour for db.t4g.micro. PostgreSQL offers richer features such as JSON columns, extensions and vector search, so it is a common default for new products. MySQL is a fine choice if your framework or team already uses it.
When should I use DynamoDB instead of SQL?
Use DynamoDB when access patterns are simple and known, scale is high, and you want no servers to manage. In Mumbai it costs $0.71 per million writes and $0.1425 per million reads on demand. Avoid it if you need flexible reporting or joins.
Is MongoDB good for a SaaS MVP?
It can be, if your data is document-shaped and changes often. For typical business data with relationships, a relational database is usually simpler long term. Managed MongoDB-compatible options include Amazon DocumentDB, at $0.10961 an hour on db.t4g.medium in Mumbai.
How much does a managed database cost in India?
Entry managed relational databases cost about $15 to $18 a month: RDS PostgreSQL db.t4g.micro in AWS Mumbai at $0.021 an hour, or Azure PostgreSQL Flexible Server B1ms in Central India at $0.0245 an hour, based on November 2026 prices.
Sources
- AWS Price List: Amazon RDS and Aurora, Asia Pacific (Mumbai), Amazon Web Services, 2026
- AWS Price List: Amazon DynamoDB, Asia Pacific (Mumbai), Amazon Web Services, 2026
- AWS Price List: Amazon DocumentDB, Asia Pacific (Mumbai), Amazon Web Services, 2026
- Azure Retail Prices API (Central India meters), Microsoft Azure, 2026
- AWS Price List: Amazon S3, Asia Pacific (Mumbai), Amazon Web Services, 2026
- AWS Activate: startup credits, Amazon Web Services, 2026
- Azure regions list, Microsoft Learn, 2026
