Lighthouse vs Netflix DGS
Lighthouse
lighthouse-php.comNetflix DGS
netflix.github.ioSide by side
| Figure | Lighthouse | Netflix DGS |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | — |
| API | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 1 |
| Spec score · GraphQL Tools | 8.1 (#7) | 7.7 (#14) |
| Purpose | Lighthouse is a framework for serving GraphQL from Laravel. | DGS is a GraphQL server framework for Spring Boot that makes it easier to create GraphQL services. |
Plans
In their own words
Lighthouse
Lighthouse is a framework for serving GraphQL from an existing Laravel application. It uses GraphQL Schema Definition Language and server-side directives to describe data and add behavior, while reusing Laravel concepts and Eloquent models. Built-in directives can create records, resolve relationships, apply query constraints, paginate results, and validate fields and arguments. Developers can add custom directives for reusable behavior, while the @can directive family can use Laravel Policies to restrict field operations. The default endpoint is `/graphql`, and any GraphQL client can use it. Lighthouse also supports standard Laravel authentication mechanisms, including Sanctum, and GraphQL subscriptions for event updates. Parsed queries can be cached with Laravel cache stores or local PHP files using OPcache; Automatic Persisted Queries are supported with Apollo compatibility. Installation uses Composer in a Laravel or Lumen project. The package is free and self-hosted. GraphQL tooling such as GraphiQL is not included, and federation support is experimental and cannot serve as a federation gateway.
Netflix DGS
Netflix DGS is a free, open-source GraphQL server framework for Spring Boot, deployed as a self-hosted API service. Its annotation-based programming model supports Java and Kotlin, with Gradle or Maven builds. DGS favors schema-first development: schema files belong in src/main/resources/schema, and its Gradle plugin can generate Java types from a GraphQL schema. A test framework runs GraphQL queries in lightweight unit tests, while data loaders batch requests to help address the N+1 problem. The framework supports GraphQL Federation, WebSocket and SSE subscriptions, file uploads, error handling, and extension points. Spring Security integration allows @Secured on data fetchers after security configuration. DGS also integrates with Spring for GraphQL, although that path currently lacks SSE subscriptions and some DGS data-loader features do not work with Spring for GraphQL data loaders. A browser GraphiQL editor and an IntelliJ plugin are included. Current documentation requires Spring Boot 3 and JDK 17; DGS 11+ with Spring Boot 4 is actively maintained, while DGS 10.x receives most feature backports until the second half of 2026.