Anvia

OpenAI-compatible APIs ​

Use @anvia/openai when a provider or gateway exposes an OpenAI-compatible HTTP API. OpenAIClient handles the compatible wire format while agents and application services continue to use Anvia model contracts.

ts
import { OpenAIClient } from '@anvia/openai'

const client = new OpenAIClient({
  apiKey: process.env.COMPATIBLE_API_KEY!,
  baseUrl: 'https://provider.example.com/v1',
})

export const model = client.completionModel({
    modelId: 'provider/model-name',
    api: "chat"
})

OpenAI-compatible does not mean OpenAI-equivalent. An endpoint may accept one Chat Completions request but differ in streaming chunks, tool calls, schemas, reasoning fields, usage, media, errors, or model discovery.

The configuration boundary ​

A compatible deployment has five important choices:

  • baseUrl selects the provider or gateway API root.
  • apiKey, or a preconfigured client, supplies server-side authentication.
  • completionModel({ modelId, api }) selects the endpoint's exact model ID and the Chat or Responses adapter.
  • headers supplies trusted gateway headers when required.

Provider-specific request fields belong in providerOptions at the narrow call site that needs them.

Select api: 'chat' for a Chat Completions-compatible endpoint. Select api: 'responses' only after proving that the endpoint implements the Responses API.

Prove compatibility per workflow ​

Do not approve an endpoint because one text request succeeded. Test every product path against the exact endpoint, adapter, model ID, account, and parameters:

  • non-streaming and streaming text
  • tools, tool choice, and multi-turn tool replay
  • schema-constrained output
  • reasoning metadata
  • image or document input
  • embeddings and media routes
  • usage reporting and failure mapping

The smallest useful integration is often completion-only. Add other model factories only after the endpoint documents and passes tests for those OpenAI-compatible routes.

Continue the setup ​

  1. Configure the endpoint and credentials.
  2. Choose Chat or Responses.
  3. Verify required capabilities.
  4. Pin models and provider parameters.
  5. Run the compatibility test plan.
  6. Review the production checklist.

Built for Anvia.