Integration

OpenAI-compatible API integration for variables and secrets

Connect Wodby apps to an inference service that implements the OpenAI API, including hosted endpoints or your own vLLM server. Supply a base URL and, when required, an API key through OPENAI_BASE_URL and OPENAI_API_KEY.

Variables

Choose your inference endpoint

Use a compatible API from services such as Modal, Together, Fireworks, or a Hugging Face Inference Endpoint, or connect a model server you operate.

Match the endpoint’s authentication

Set the API base URL, usually ending in /v1. The API key is optional for endpoints that do not require authentication.

Supported workflows

How OpenAI-compatible API fits into Wodby

Each supported integration kind maps OpenAI-compatible API into a specific part of the delivery workflow, whether that means pulling source code, running pipelines, relaying email, connecting managed services, or injecting provider configuration into workloads.

Variables

API

Configure apps to call a hosted or self-managed OpenAI-compatible inference API through environment variables.

  • Set the API base URL expected by your client, usually ending in /v1.
  • Provide an API key if the endpoint requires authentication; otherwise the integration’s key field can be left empty.
  • Connected services receive OPENAI_BASE_URL and OPENAI_API_KEY. The endpoint must support the API operations your app uses.

Configuration surfaced to workloads

OPENAI_BASE_URLOPENAI_API_KEY
See integration workflows

Related features

Explore the surrounding workflow

This provider usually sits next to stack-based delivery, backup workflows, managed databases, or integration-backed configuration. These pages cover the product surfaces most relevant to OpenAI-compatible API.

See integration workflows

Store provider keys, tokens, and DSNs once, then reuse them across environments.

View page

Browse all integrations

Compare this provider with the rest of the integration catalog and other supported workflows.

View page

Stack fit

Stacks that commonly pair with OpenAI-compatible API

Wodby does not treat integrations as isolated setup screens. They become part of stack delivery, so the strongest matches are the stacks where this provider solves a repeated operational problem.

  • Python

    Stack

    Python application stack for custom web apps, APIs, workers, and automation services.

    Configure Python clients to call your chosen compatible inference endpoint using runtime variables.

    View stack

  • Next.js

    Stack

    Next.js application stack for frontend and full-stack projects that need a reusable deployment setup.

    Next.js apps can receive provider configuration through the services that need it instead of scattering values across the repository.

    View stack

  • Node.js

    Stack

    Node.js application stack for APIs, workers, and custom services with a repeatable runtime setup.

    Node.js services can receive provider settings as runtime variables instead of repo-local env files.

    View stack

  • Laravel

    Stack

    Laravel application stack for teams that want a reusable setup for web, worker, and queue-based PHP projects.

    Laravel is a strong fit when provider credentials and configuration should stay centralized at the stack layer.

    View stack

Compare providers or keep reading

Use the catalog to compare other providers with the same integration kind, or read the broader feature pages to understand how integrations fit into Wodby stacks, CI/CD, backups, and managed services.

Browse all integrationsIntegration feature overview