OpenAI-Compatible Adapter

A generic adapter for any provider that exposes an OpenAI-compatible Chat Completions API. Works out of the box with OpenAI, Groq, Together AI, DeepSeek, Fireworks, Mistral, Perplexity, Ollama (local), and any other server that speaks the same wire format.

Install

npm install @flint/adapter-openai-compat

Usage

OpenAI

import { openaiCompatAdapter } from '@flint/adapter-openai-compat';
import { call } from 'flint';

const adapter = openaiCompatAdapter({
  apiKey: process.env.OPENAI_API_KEY,
  baseUrl: 'https://api.openai.com/v1',
});

const result = await call({
  adapter,
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Hello!' }],
});

if (result.ok) {
  console.log(result.value.message.content);
}

Groq

import { openaiCompatAdapter } from '@flint/adapter-openai-compat';
import { call } from 'flint';

const adapter = openaiCompatAdapter({
  apiKey: process.env.GROQ_API_KEY,
  baseUrl: 'https://api.groq.com/openai/v1',
});

const result = await call({
  adapter,
  model: 'llama-3.3-70b-versatile',
  messages: [{ role: 'user', content: 'Hello!' }],
});

if (result.ok) {
  console.log(result.value.message.content);
}

Ollama (local)

import { openaiCompatAdapter } from '@flint/adapter-openai-compat';
import { call } from 'flint';

const adapter = openaiCompatAdapter({
  // No apiKey required for local Ollama
  baseUrl: 'http://localhost:11434/v1',
});

const result = await call({
  adapter,
  model: 'llama3.2',
  messages: [{ role: 'user', content: 'Hello!' }],
});

if (result.ok) {
  console.log(result.value.message.content);
}

Constructor Options

Option	Type	Required	Description
baseUrl	string	Yes	Base URL of the OpenAI-compatible endpoint (no trailing slash)
apiKey	string	No	Bearer token sent as Authorization: Bearer <key>. Omit for local servers that do not require auth
defaultHeaders	Record<string, string>	No	Extra headers merged into every request — useful for provider-specific auth schemes or versioning headers
fetch	typeof globalThis.fetch	No	Custom fetch implementation

Capabilities

Capability	Supported
Streaming	Yes
Tool use	Yes
Parallel tools	Yes
Prompt caching	No
Structured output	Yes
Vision (images)	Provider-dependent
Token counting	Provider-dependent

Prompt caching is not supported because it is an Anthropic-specific feature. Individual providers may offer their own caching mechanisms; consult their docs.

Zero Runtime Dependencies

@flint/adapter-openai-compat has no npm runtime dependencies. It uses the platform's native fetch (or a custom implementation you supply via the fetch option) and types re-exported from the core flint package.

openAICompatAdapter() options

function openAICompatAdapter(options: {
  apiKey: string;
  baseURL: string;
  defaultHeaders?: Record<string, string>;
}): ProviderAdapter

Provider-specific configurations

OpenAI:

openAICompatAdapter({ apiKey: process.env.OPENAI_API_KEY!, baseURL: 'https://api.openai.com/v1' })

Groq (fast inference):

openAICompatAdapter({ apiKey: process.env.GROQ_API_KEY!, baseURL: 'https://api.groq.com/openai/v1' })
// Note: Groq has aggressive rate limits on free tier

Together AI:

openAICompatAdapter({ apiKey: process.env.TOGETHER_API_KEY!, baseURL: 'https://api.together.xyz/v1' })

DeepSeek:

openAICompatAdapter({ apiKey: process.env.DEEPSEEK_API_KEY!, baseURL: 'https://api.deepseek.com/v1' })

Ollama (local):

openAICompatAdapter({ apiKey: 'ollama', baseURL: 'http://localhost:11434/v1' })
// apiKey is required by the adapter but not validated by Ollama

Capabilities

adapter.capabilities = {
  streaming: true,
  toolCalling: true,   // supported by most OpenAI-compat providers
  vision: false,       // varies by provider
  cost: false,         // most providers don't report cost
  promptCaching: false,
};

Cost and caching are not reported — maxDollars budgets won't trigger with this adapter.

See Also

• Anthropic Adapter
• Writing an Adapter
• FAQ: Does Flint support local models?
• Source: packages/adapter-openai-compat/src/index.ts