@agentplat/model-openai-compatible

A dependency-light ModelAdapter for servers that implement the OpenAI Chat Completions wire format. It uses the platform fetch implementation and does not install an OpenAI SDK.

Installation (developer preview)

Install the coordinated preview explicitly:

npm install @agentplat/model-openai-compatible@next

Keep all @agentplat/* packages on the same release version. npm's default latest tag can point to an older preview. See the release channels for distribution status and version selection.

import { chatModel } from "@agentplat/model-openai-compatible";

const model = chatModel({
  provider: "openai",
  apiKey: process.env.OPENAI_API_KEY,
  defaultModel: process.env.OPENAI_MODEL,
});

chatModel has SDK-free presets for openai, gemini, ollama and openrouter. Use compatible with an explicit baseURL for another Chat-Completions server:

const localModel = chatModel({
  provider: "compatible",
  baseURL: "http://localhost:1234/v1",
  requireApiKey: false,
  defaultModel: "local-model",
});

The adapter does not read environment variables. Provider capabilities and parameter support vary. For a provider with a non-compatible HTTP protocol, implement the small public ModelAdapter interface from @agentplat/model. Use @agentplat/provider-openai when an application needs the OpenAI Agents SDK runner and the provider-specific capabilities implemented by that adapter.