DreamFactory

Self-hosted API generation, honestly reviewed. Useful technology buried under enterprise pricing you can’t see.

TL;DR

• What it is: Open-source (Apache 2.0) platform that connects to any database and auto-generates a full REST API with RBAC, OpenAPI docs, and authentication — no backend code required [README][5].
• Who it’s for: Developers and technical teams who need to expose internal databases as APIs, connect AI agents to enterprise data, or modernize legacy systems without writing CRUD boilerplate. Not for non-technical founders looking for a drag-and-drop experience [README][4].
• Cost savings: Eliminating hand-coded API development. DreamFactory’s own calculator claims $45,719 saved per API and $201,783 in annual dev cost savings [website]. These are marketing numbers — treat them skeptically — but the underlying time argument is real.
• Key strength: Genuine breadth of database support. MySQL, PostgreSQL, SQL Server, MongoDB, Oracle, Snowflake, Cassandra, IBM DB2, Azure Blob, and 20+ more, all through one unified API layer [README][website].
• Key weakness: The open-source version is deliberately crippled. The website explicitly states: “The open-source version offers limited features. For the full experience, including access to all databases, advanced customization options, & robust security features — please book a demo with our engineers.” [website] There is no public pricing. You will fill out a sales form.

---

What is DreamFactory

DreamFactory connects to your existing database, reads the schema, and generates a complete REST API with full CRUD operations, relationship traversal, stored procedure access, and live OpenAPI/Swagger documentation — without you writing a single line of backend code. It’s built on the Laravel framework and has been around since 2013, when it was released on GitHub under the Apache 2.0 license [2].

The core use case is straightforward: you have a database, you need an API, and writing that API by hand takes a senior engineer weeks of work per data source. DreamFactory compresses that to minutes [5]. Once connected, every table becomes a REST endpoint. You get filtering, sorting, pagination, and relationship handling out of the box.

Beyond basic CRUD, DreamFactory has more recently positioned itself as an AI data gateway: the idea that your internal databases should be queryable by AI agents, local LLMs, and tools like Claude or ChatGPT, but without those agents ever touching database credentials directly [README]. Every API it generates can be exposed as an MCP (Model Context Protocol) server, meaning Claude Desktop, Cursor, or any MCP-compatible tool can query your database through governed, permission-scoped endpoints [README].

The client list on their website includes Intel, McAfee, Nike, McKesson, ExxonMobil, and NIH — mostly large enterprises with compliance requirements and existing database estates that need API access without ripping out their infrastructure [website]. That audience tells you something important about who this is actually built for.

---

Why people choose it

The case for DreamFactory appears in three recurring situations:

The legacy database modernization problem. You have a database that was built in 2003, it runs the business, and you need modern applications to talk to it. Nobody wants to write a PHP SOAP bridge in 2026. DreamFactory wraps the old schema in a REST API with OAuth and RBAC, and the frontend team never needs to know what’s underneath [5][website]. One case study describes the Vermont Department of Transportation using DreamFactory to connect 1970s-era systems with modern applications through REST APIs — exactly the kind of thing that would otherwise require a full replatforming project [website].

The AI data access problem. LLMs are useful for internal tooling, but you can’t give your AI agent a raw database connection. DreamFactory lets you expose specific tables and views to an AI agent with field-level security, rate limiting, and audit logging [README]. The agent sees governed API endpoints, not credentials or schemas. The NIH case study mentions using this pattern to accelerate grant application analytics without rebuilding their data infrastructure [website].

The “I need an API but I’m not an API team” problem. SoftwareAdvice reviewers describe implementing DreamFactory to integrate Salesforce with an ERP that had no API, citing limited budget as the reason they needed a shortcut rather than building middleware [4]. Another reviewer calls out the Salesforce integration specifically and describes the setup as something that would “normally take months” done in hours [website customer quote].

What it is not is a Zapier replacement, a no-code workflow builder, or something a non-technical founder can configure alone. The Getting Started Guide assumes you know what REST, HTTP methods, and RBAC mean [5]. SoftwareAdvice rates ease of use at 4.0/5 from 11 reviews — adequate, but this is an enterprise tool being evaluated by enterprise users. For a non-technical founder starting from scratch, the learning curve is steep and the open-source limitations make the free tier essentially a demo.

---

Features

Core API generation:

• Point at any supported database, DreamFactory introspects the schema and generates full CRUD endpoints immediately [README][5]
• Supports GET (with filtering, sorting, pagination, related records), POST, PUT, PATCH, DELETE [5]
• Stored procedure access — call existing business logic through REST endpoints [4][5]
• Auto-generated OpenAPI/Swagger documentation that stays current with schema changes [README][website]
• Webhooks and event-based triggers [README]

Security and access control:

• Role-based access control down to the field and row level — different API consumers see different data [README][4]
• API keys, OAuth 2.0, SSO, LDAP [README]
• Identity passthrough — credentials stay inside DreamFactory, never exposed to consuming applications [README][website]
• Rate limiting per service, role, or user [README]
• Audit logging [README]

AI and LLM integration:

• MCP server mode — auto-expose any API service as an MCP endpoint for Claude Desktop, Cursor, Windsurf, or any MCP-compatible client [README]
• The MCP flow: LLM queries → DreamFactory MCP endpoint → RBAC check → database → governed response [README]
• REST APIs compatible with any AI agent that can make HTTP calls [README]

Supported data sources (20+ connectors): Relational: MySQL, PostgreSQL, SQL Server, Oracle, SQLite, IBM DB2, Snowflake [README][website] NoSQL: MongoDB, Cassandra, DynamoDB [README] File/object storage: Amazon S3, Azure Blob [website] SaaS: Salesforce [4][website]

Deployment:

• Docker Compose (quickest path) [README]
• Helm charts for Kubernetes [README]
• Linux installer scripts for bare-metal [README]
• Windows installer available [README]

What’s gated behind commercial licensing: The website is vague here, but states the open-source version has “limited features” versus the commercial version. Based on what’s publicly documented, the commercial/enterprise tier adds: full database connector breadth, advanced customization, “robust security features,” and dedicated support [website]. The exact split between open-source and commercial is deliberately not documented in detail — which is itself a data point.

---

Pricing: SaaS vs self-hosted math

This section is normally where the savings math lives. For DreamFactory, that math is hidden.

DreamFactory’s pricing model: Not publicly available. The website says “pricing available upon request” and routes you to a sales call or a demo [4][website]. There is no published tier structure, no listed monthly fee, and no self-serve purchase option.

What this means in practice: If you’re a solo founder or small team, this is a red flag. “Contact sales” pricing reliably signals either (a) the cost is significant and they don’t want to scare you away before the demo, or (b) pricing is heavily customized by company size and usage, which usually means small accounts don’t get favorable terms.

For comparison, the self-hosted software tools that compete in adjacent spaces:

• PostgREST (open source, Apache 2.0): free, generates PostgreSQL REST APIs — no commercial upsell, no gated features [open source]
• Hasura Community (open source): free for self-hosted, with commercial add-ons clearly listed starting around $299/mo [public pricing]
• Directus (open source, BSL): free self-hosted community edition, clear cloud pricing [public pricing]

DreamFactory’s 2013 blog announcement [2] positioned it as Apache 2.0 open source solving a problem for developers. The 2026 version of the product has evolved into an enterprise platform where the Apache 2.0 repository is essentially a trial, and the actual product is commercial. That’s a legitimate business model — but know what you’re walking into.

The self-hosted cost floor: Running the open-source version on a VPS costs $5–20/month in infrastructure. Whether that’s sufficient for your use case depends entirely on which features end up behind the commercial gate — which you’ll only find out during a sales demo.

---

Deployment reality check

The technical setup is legitimately well-documented. DreamFactory provides a Docker Compose path, Helm charts, Linux installer scripts, and Windows installer — more than most self-hosted tools offer [README]. The Getting Started Guide at guide.dreamfactory.com is thorough [5].

What you need:

• A Linux server (or Docker on any OS) with at least 2–4GB RAM
• Docker and docker-compose, or a Laravel/PHP environment for manual install
• The database(s) you want to expose (DreamFactory connects externally, doesn’t need to host them)
• An SMTP provider if you want email notifications
• A reverse proxy (nginx or Caddy) for production HTTPS

What can go sideways:

• The open-source feature ceiling is unclear until you hit it. You may spend time setting up the platform only to discover the database connector you need or the security feature your use case requires is commercial-only [website].
• SoftwareAdvice reviewers [4] give customer support 4.0/5 — acceptable, but several note that documentation and support quality drop off significantly on the open-source tier compared to commercial accounts.
• Stored procedure support, while listed as a feature, drew a specific complaint from one reviewer: “run processes using SQL stored procedures which gave us a bit of a headache at the beginning because of lack of full support” [4].
• DreamFactory is PHP/Laravel, not a stateless binary. Memory and performance tuning for high-traffic API workloads requires PHP server expertise that some teams won’t have.
• The GitHub repository has 1,755 stars — modest for a project that’s been around since 2013. The community is smaller than alternatives like PostgREST or Hasura, which affects how much StackOverflow-style help you’ll find for edge cases.

Realistic setup time for a developer: 1–3 hours from zero to working API on Docker. For non-technical founders: this is not the right tool.

---

Pros and Cons

Pros

• Broad database support out of the box. MySQL, PostgreSQL, SQL Server, Oracle, MongoDB, Snowflake, Cassandra, IBM DB2, S3, Salesforce, and more through a single interface [README][website]. Competing open-source tools often cover only one database type well.
• MCP-native AI data gateway. The architecture of exposing governed REST APIs as MCP endpoints is the correct mental model for connecting AI agents to enterprise data in 2026. Credentials never leave DreamFactory [README].
• Legitimately saves API development time. Customer quotes across G2 and SoftwareAdvice consistently describe days or weeks of backend work compressed to minutes [4][website]. For teams that need internal APIs quickly, this is real.
• Enterprise pedigree. ExxonMobil, NIH, Deloitte, McKesson, Nike — these aren’t vanity logos. DreamFactory has been in production at compliance-heavy organizations for years [website]. If you need a self-hosted API layer that’s going to stay maintained, that track record matters.
• Apache 2.0 license on the repo. You can fork it, audit it, and run it without a vendor agreement — at least the open-source portion [README].
• Full OpenAPI/Swagger docs auto-generated and kept current with schema changes. This alone would take a developer weeks to write manually [README][5].
• G2 rating of 4.7 and SoftwareAdvice at 4.1/5 across 11 reviews — not deep review volume, but the reviews that exist are consistently positive from people in production [4][website].

Cons

• No public pricing. “Contact sales” means this product is not built for small teams or solo founders. The sales cycle is required before you know if you can afford it [4][website].
• Open-source version is a demo. The website explicitly tells you the open-source version has “limited features” and pushes you toward a sales call for the full experience [website]. The Apache 2.0 license on GitHub doesn’t tell you what’s actually locked.
• 1,755 GitHub stars after 13 years is a small community. PostgREST has 25K+ stars, Hasura has 31K+. A smaller community means fewer extensions, fewer community guides, and fewer engineers who know the tool when you’re hiring [README metadata].
• Stored procedure support has documented rough edges. Real user complaint about setup friction for SQL stored procedures [4].
• PHP/Laravel runtime adds dependency overhead and requires PHP expertise for performance tuning, which many modern teams don’t have.
• Not for non-technical users. The tool requires understanding of REST APIs, RBAC, database schemas, and deployment. The Getting Started Guide assumes substantial technical background [5].
• AI/LLM positioning feels early. The MCP integration is real and technically sound, but the marketing around “let AI query your data” is ahead of most teams’ actual AI workflows. Don’t buy this primarily for the AI angle.

---

Who should use this / who shouldn’t

Use DreamFactory if:

• You have an existing database (legacy or modern) and need REST API access for multiple applications or teams without building it by hand.
• You’re a developer or technical team lead who needs to ship internal APIs fast and can navigate a sales conversation to understand pricing.
• You’re connecting AI agents or local LLMs to internal databases and need a governed access layer with field-level security.
• You’re at a company large enough to justify an enterprise tool conversation — think 50+ employees, compliance requirements, or multiple databases to expose.
• You need to support legacy data sources like IBM DB2, Oracle, or mainframe systems that other tools don’t cover.

Skip it (use PostgREST instead) if:

• Your database is PostgreSQL and you want a free, fully open-source, no-sales-call solution with 25K+ GitHub stars and deep community documentation. PostgREST does exactly what DreamFactory does for PostgreSQL, forever free.

Skip it (use Hasura instead) if:

• You want GraphQL instead of REST, or you want a tool with transparent commercial pricing and a larger community. Hasura’s open-source version has a clear feature boundary and the commercial tier is publicly priced.

Skip it (use Directus instead) if:

• You need a full data platform with API generation plus a content management UI and app builder. Directus covers more ground for teams building internal tools or headless CMS use cases.

Skip it entirely if:

• You’re a non-technical founder. This requires a developer to set up, configure, and maintain. No amount of marketing language about “no code required” changes what’s actually required to operate it safely in production.

---

Alternatives worth considering

• PostgREST — free, fully open-source, excellent for PostgreSQL specifically. No vendor relationship, no feature gating, massive community. If PostgreSQL is your database, start here.
• Hasura — GraphQL-first API generation over PostgreSQL, MySQL, and others. Open-source community edition with transparent commercial pricing. More modern developer experience.
• Directus — REST and GraphQL API generation over SQL databases with a full admin UI. Business Source License, but the self-hosted version is widely used and free for most cases.
• Supabase — full backend platform (PostgreSQL + auto-generated REST and GraphQL APIs + auth). Higher-level abstraction but dramatically better developer experience for new projects.
• Kong — API gateway rather than API generator. Use if you already have APIs and need routing, rate-limiting, and auth management across them [3].
• Apigee (Google Cloud) — the enterprise incumbent DreamFactory competes against directly [3]. More features, more complex, more expensive, fully managed. For large enterprises already in Google Cloud.

---

Bottom line

DreamFactory solves a real problem: you have databases, you need APIs, and writing CRUD boilerplate by hand is expensive. For enterprises with legacy data estates, compliance requirements, and the budget to have a sales conversation, it’s a credible choice with a decade-long enterprise track record and genuine database breadth. The MCP integration is a thoughtful response to the AI agent access problem.

But the open-source bait-and-switch is real. If you’re evaluating DreamFactory because it showed up on a self-hosted tools list and the Apache 2.0 license looked promising — read the website copy carefully before investing setup time. The open-source repository is a limited edition of the commercial product, and the commercial product has no public pricing. For solo founders, small teams, or anyone who wants a genuine open-source API generator without a sales call, PostgREST, Hasura, or Directus will serve you better. If you have the scale to justify enterprise tooling and a compliance case for self-hosted infrastructure, DreamFactory belongs on your shortlist.

---

Sources

1. SourceForge — DreamFactory Reviews. https://sourceforge.net/software/product/DreamFactory/
2. Terence Bennett, DreamFactory Blog — “DreamFactory Enterprise: DreamFactory Management Made Easy” (July 6, 2018). https://blog.dreamfactory.com/dreamfactory-enterprise-dreamfactory-management-made-easy
3. Spencer Nguyen, DreamFactory Blog — “The Top 9 Apigee Alternatives” (January 11, 2026). https://blog.dreamfactory.com/the-best-apigee-alternatives
4. SoftwareAdvice — “DreamFactory Software Reviews, Demo & Pricing - 2026” (4.1/5, 11 reviews). https://www.softwareadvice.com/bi/dreamfactory-profile/
5. DreamFactory Getting Started Guide — “DreamFactory REST API Platform: Overview & Key Concepts”. https://guide.dreamfactory.com/docs/introducing-rest-and-dreamfactory/

Primary sources:

• GitHub repository: https://github.com/dreamfactorysoftware/dreamfactory (1,755 stars, Apache-2.0 license)
• Official website: https://www.dreamfactory.com
• Documentation: https://docs.dreamfactory.com