Data API Builder Alternative

Microsoft's Data API Builder Is a Free Engine.
DreamFactory Is the Platform.

Data API builder is a free, MIT‑licensed engine that turns SQL Server, PostgreSQL, MySQL, and Azure databases into REST, GraphQL, and MCP endpoints from one JSON file. DreamFactory generates the governed REST API — and MCP tools for AI agents — from your database and adds an admin console, scripting, rate limits, audit logs, 30+ connectors, and published support.

Book a Demo Explore DreamFactory Features
What the Docs Say

Five Things to Weigh Before Standardizing on Data API Builder

Data API builder is free under the MIT license, maintained by Microsoft, and good at its job: if you need to expose SQL Server as a REST API without writing code, it does that from a single configuration file. The questions worth asking are about where its documented design stops, and what your team would build or buy around it.

1. The Database List Is SQL Server, PostgreSQL, MySQL, and Cosmos DB

As documented, DAB supports SQL Server 2016+, Azure SQL, SQL database in Fabric, PostgreSQL 11+, MySQL 8+, Synapse dedicated SQL pools, and Azure Cosmos DB for NoSQL through GraphQL and MCP (no REST; Cosmos DB has a native REST API). Oracle, Snowflake, IBM Db2, MongoDB, and SAP HANA are not in the supported list, and cross‑source relationships are not supported.

2. Identity Is Delegated, Not Owned

DAB validates tokens from Entra ID, App Service authentication, or any OIDC/JWT issuer (JWT‑bearing clients authenticate directly), and the X-MS-API-ROLE header selects each request's role, matched against the token's roles claim. DAB does not issue API keys (its feature matrix: not supported) or manage user accounts; for key‑based clients, Microsoft's MCP guidance is a gateway such as Azure API Management.

3. Rate Limiting Is Not in the Config Schema

The published configuration schema (current release 2.0.12) has no rate‑limit, quota, or client API‑key property; Microsoft's feature matrix marks throttling as supported for GraphQL but documents no setting for it. Documented request shaping is page‑size caps, GraphQL depth limit, maximum response size, and level‑1/level‑2 caching; Microsoft's production security best-practices list for its MCP server includes “Enable rate limiting” as a separate step.

4. Configuration Is a File and a CLI

Configuration is dab-config.json and the dab CLI, with an integrated Data API builder UI (with GitHub Copilot chat) inside the MSSQL extension for VS Code and SQL Server autoentities patterns that expose matching tables without declaring each one. Roles and permissions live in that file; in production mode, hot reload covers the log level and other changes need a restart.

5. Support Runs Through GitHub Issues, Without an SLA

Project engineers review GitHub issues periodically; Microsoft's support statement says the review period is not set by an SLA and resolution is not guaranteed, while Azure support tickets cover the hosting service or database underneath. Like any self‑hosted container, DAB does not update itself; watching releases and applying updates stays with your team. Weigh that against the support you expect.

The DreamFactory Way

The Generated API, Plus the Platform to Run It

DreamFactory generates the documented REST API from your schema in minutes and ships the operational layer around it: a console for users, roles, and keys, scripting, rate limits, audit logs, 30+ connectors, and published support packages.

Instant REST API Generation

Connect SQL Server, MySQL, PostgreSQL, Oracle, MongoDB, Snowflake, and more — DreamFactory auto‑generates a full, documented REST API in minutes, with Swagger/OpenAPI docs included.

See all connectors →

Users, Roles, and Keys in One Console

Role‑based access control and API keys come standard on every generated endpoint, with OAuth/JWT sign‑in supported, and the DF Linux Professional tier and above add rate limiting and audit logging — all managed in a web admin console, with every admin action also exposed as a REST API.

Explore features →

Your Identity Provider, Mapped to Roles

SAML 2.0, OpenID Connect, OAuth 2.0, LDAP, and Active Directory sign users in (from the DF Linux Professional tier), and IdP groups or attributes map to DreamFactory roles automatically at login — including Azure AD group IDs through OIDC — with identity passthrough so the database's own audit log shows the real user.

Logic in Languages You Know

Business rules run server‑side in Node.js, Python, or PHP on any endpoint — as pre‑process, post‑process, or queued scripts, or as standalone scripted services — ordinary application code, no restart required.

See a scripted AI workflow on SQL Server →

Support and Maintenance, Published

The core is open source under Apache 2.0. Commercial tiers include the SQL Server connector from the $1,500/month DF Linux Lite tier, with Oracle, Snowflake, and Db2 among unlimited connectors from the $4,000/month DF Linux Professional tier (all with unlimited roles, keys, and API requests), and published support and maintenance packages extend to a 2‑hour SLA with a dedicated engineer.

SQL Server connector →

Ready for AI Agents

A built‑in MCP server exposes your databases as governed tools for MCP clients such as Claude, ChatGPT, Cursor, and VS Code — with the same RBAC and audit trail as every other call.

Side by Side

DreamFactory vs Data API Builder at a Glance

DreamFactory is a self‑hosted Data API builder alternative for teams whose priority is governed REST APIs — and MCP access for AI agents — on their databases, generated in minutes, managed from a console, with published pricing. Data API builder is Microsoft's free, MIT‑licensed engine that serves REST, GraphQL, and MCP endpoints from a JSON configuration.

CapabilityWhy it mattersDreamFactoryData API builder
REST APIs from a databaseThe faster the data layer exists, the sooner apps and AI can use itAuto‑generated in minutes with OpenAPI docs, from 30+ connectors including SQL Server, PostgreSQL, MySQL, Oracle, Snowflake, and MongoDBOne JSON configuration, OpenAPI 3.0.1, for SQL Server, Azure SQL, PostgreSQL, MySQL, and Synapse dedicated pools; Cosmos DB for NoSQL via GraphQL and MCP (it has a native REST API)
Endpoint typesClients and agents differ in what they speakREST with live OpenAPI docs — filters, pagination, sorting, related‑table joins, and virtual foreign keys — plus MCP tools from the same governed servicesREST, GraphQL, and MCP served simultaneously from one configuration (MCP since version 1.7), including GraphQL relationships and aggregation on SQL Server
Access managementSomeone has to issue keys and assign rolesAdmin console and System API for users, roles, and API keys; SAML, OIDC, OAuth, and LDAP/AD with IdP group‑to‑role mappingDelegated to Entra ID, App Service authentication, or any OIDC/JWT issuer; role via the X-MS-API-ROLE header; key‑based clients use a gateway such as Azure API Management, per Microsoft's MCP guidance
Rate limits & auditProduction APIs need throttles and a record of who called whatLimits per user, role, service, or endpoint (HTTP 429), and per‑call audit logging with real user identity through Logstash/ELKOpenTelemetry, Application Insights, Log Analytics, and file sinks; On‑Behalf‑Of and SESSION_CONTEXT pass caller identity to SQL Server and Azure SQL; no rate‑limit or quota setting in the published schema (feature matrix: GraphQL throttling, no documented setting)
Custom logicReal APIs transform, validate, and enforce rulesNode.js, Python, or PHP scripts on any endpoint — pre‑process, post‑process, or queued — plus standalone scripted services; stored procedures, functions, and views also exposed as REST endpointsStored procedures as REST/GraphQL entities and custom MCP tools on SQL Server and Synapse dedicated pools (not PostgreSQL or MySQL, as documented); per‑role database policies filter rows by token claims
Deployment & supportRegulated data often can't leave your network, and production needs a support pathSelf‑hosted on Linux, Windows Server, Docker, or Kubernetes, air‑gapped included; published support packages up to a 2‑hour SLA with a dedicated engineerStrong parity on hosting: official x86‑64 container image, on‑premises, any cloud, or air‑gapped; GitHub‑issue support without an SLA, as documented (Azure tickets cover the hosting service and database); image updates applied by your team

Competitor details verified September 2026 from vendor documentation. Product names are trademarks of their respective owners.

An Honest Framing

Which One Fits Your Team?

Choose Data API builder if…

Your databases are SQL Server, Azure SQL, PostgreSQL, MySQL, or Azure Cosmos DB for NoSQL; your identity already lives in Entra ID or another JWT issuer; you want REST, GraphQL, and MCP tools from one configuration; and your team is set up to own a config‑driven container (plus a gateway if clients need API keys) and its release cadence — with no license cost.

Choose DreamFactory if…

Your goal is secure, documented REST APIs — and MCP access for AI agents — on the databases you already run, whether that is SQL Server alone or SQL Server alongside Oracle, Snowflake, Db2, and MongoDB. You want a console for users, roles, and keys, scripting in languages your team knows, the entire platform self‑hosted (air‑gapped included), published support packages behind it, and a working API this week.

Frequently asked questions

What is the difference between DreamFactory and Data API builder?

They do the same core job at different scopes. Data API builder is Microsoft's free, MIT-licensed engine: one JSON configuration file turns SQL Server, Azure SQL, PostgreSQL, and MySQL objects into REST, GraphQL, and MCP endpoints (Azure Cosmos DB for NoSQL via GraphQL and MCP), with authentication delegated to Entra ID or any other JWT issuer, and a gateway or proxy such as Azure API Management in front when a client needs API keys. DreamFactory is a self-hosted API generation platform: it produces documented, governed REST endpoints — and MCP tools for AI agents — from a database in minutes, with an admin console, API keys, RBAC, rate limiting, audit logging, and scripting built in, across 30+ connectors including Oracle, Snowflake, Db2, and MongoDB.

How much does Data API builder cost compared to DreamFactory?

Data API builder is free: open source under the MIT license, and Microsoft's documentation states it is free to use in any cloud or on-premises. Hosting, updates, and any gateway in front of it are yours to run, and support is through GitHub issues without an SLA (Azure support tickets cover the hosting service and database). DreamFactory has an Apache-licensed open-source edition, and its Linux commercial tiers carry published flat monthly prices and include the SQL Server connector, the admin console, and unlimited roles, keys, API creation, and API requests: from $1,500/month for a single connector of your choice and $4,000/month for unlimited connectors (billed annually). Support and maintenance packages — up to a 2-hour SLA with a dedicated engineer — are published on the same pricing page.

Should I use Azure API Management instead?

They are different layers. Azure API Management is Microsoft's managed API gateway for publishing and securing APIs that already exist; its v2 tiers are billed hourly per unit (about $0.21 to $3.84 per hour at East US list prices in September 2026) plus per-call charges above the included volume on Basic v2 and Standard v2, with a Consumption tier at $0.035 per 10,000 calls after the first million. It governs an API you already have rather than generating one from a database — Microsoft's Data API builder MCP guidance points to a gateway or proxy such as API Management when a client needs key-based access. DreamFactory generates the API and includes keys, roles, rate limits, and audit logging, and can sit behind API Management like any other backend.

Can Microsoft Fabric expose my SQL Server as an API?

Not directly for an on-premises SQL Server. Fabric's API for GraphQL exposes GraphQL over Fabric items — Data Warehouse, SQL database in Fabric, Lakehouse SQL analytics endpoints, mirrored databases, and Azure SQL Database — and Fabric data agents answer read-only questions over Fabric items on a paid F2-or-higher Fabric capacity (or Power BI Premium P1 or higher); an on-premises SQL Server reaches Fabric through mirroring with an on-premises or virtual network data gateway. For a REST API on the SQL Server you already run, in place, both Data API builder and DreamFactory are built for that job — DreamFactory with the console, keys, roles, and audit trail included.

What about sp_invoke_external_rest_endpoint, PostgREST, or Supabase for SQL Server?

sp_invoke_external_rest_endpoint runs in the other direction: it lets SQL Server 2025, Azure SQL Database, Azure SQL Managed Instance, and SQL database in Fabric call an outbound HTTPS endpoint from T-SQL (disabled by default on SQL Server 2025) — it does not expose your database as an API. PostgREST serves a REST API from an existing PostgreSQL database, and Supabase provides its REST API through PostgREST on top of Postgres, so neither is a SQL Server option. For SQL Server, Data API builder is the free, configuration-driven engine, and DreamFactory is the self-hosted platform that generates the API and governs it with a console, keys, roles, rate limits, scripting, and audit logs.

Can AI agents use DreamFactory APIs?

Yes. DreamFactory includes a built-in MCP (Model Context Protocol) server that exposes your databases as standardized tools for MCP-compatible clients such as Claude, ChatGPT, Cursor, and VS Code — governed by the same role-based access control and audit logging as every other API call.

Free 30-Minute Demo

See SQL Server Become a Governed REST API

Give one of our engineers 30 minutes and watch a secure, documented REST API come to life on your data — console, keys, roles, and audit trail included.

Book a Demo Read the Docs