OpenAI Compatible API plugin for Genkit

@genkit-ai/compat-oai is a plugin for using OpenAI Compatible APIs with Genkit.

This Genkit plugin allows to use OpenAI models through their official APIs.

Official OpenAI-compatible provider documentation:

• OpenAI
• xAI (Grok)
• DeepSeek
• Other compatible APIs

Supported models

This plugin also supports OpenAI models, and custom models from other model providers.

Installation

Install the plugin in your project with your favorite package manager:

• npm install @genkit-ai/compat-oai
• yarn add @genkit-ai/compat-oai
• pnpm add @genkit-ai/compat-oai

Usage

Initialize

import dotenv from 'dotenv';
import { genkit } from 'genkit';
import openAI, { gpt35Turbo } from '@genkit-ai/compat-oai';

dotenv.config();

const ai = genkit({
  plugins: [openAI({ apiKey: process.env.OPENAI_API_KEY })],
  // specify a default model if not provided in generate params:
  model: gpt35Turbo,
});

Basic examples

The simplest way to generate text is by using the generate method:

const response = await ai.generate({
  model: gpt4o
  prompt: 'Tell me a joke.',
});

console.log(response.text);

Multi-modal prompt

const response = await ai.generate({
  model: gpt4o,
  prompt: [
    { text: 'What animal is in the photo?' },
    { media: { url: imageUrl } },
  ],
  config: {
    // control of the level of visual detail when processing image embeddings
    // Low detail level also decreases the token usage
    visualDetailLevel: 'low',
  },
});
console.log(response.text);

Text Embeddings

import { textEmbeddingAda002 } from '@genkit-ai/compat-oai';

const embedding = await ai.embed({
  embedder: textEmbeddingAda002,
  content: 'Hello world',
});

console.log(embedding);

Within a flow

import { z } from 'genkit';

export const jokeFlow = ai.defineFlow(
  {
    name: 'jokeFlow',
    inputSchema: z.string(),
    outputSchema: z.string(),
  },
  async (subject) => {
    const llmResponse = await ai.generate({
      prompt: `tell me a joke about ${subject}`,
    });
    return llmResponse.text;
  }
);

Tool use

import { z } from 'genkit';

// ...initialize genkit (as shown above)

const createReminder = ai.defineTool(
  {
    name: 'createReminder',
    description: 'Use this to create reminders for things in the future',
    inputSchema: z.object({
      time: z
        .string()
        .describe('ISO timestamp string, e.g. 2024-04-03T12:23:00Z'),
      reminder: z.string().describe('the content of the reminder'),
    }),
    outputSchema: z.number().describe('the ID of the created reminder'),
  },
  (reminder) => Promise.resolve(3)
);

const result = await ai.generate({
  tools: [createReminder],
  prompt: `
  You are a reminder assistant.
  If you create a reminder, describe in text the reminder you created as a response.

  Query: I have a meeting with Anna at 3 for dinner - can you set a reminder for the time?
  `,
});

console.log(result.text);

Custom models & other Cloud providers

import { GenerationCommonConfigSchema, genkit, z } from 'genkit';
import { ModelInfo } from 'genkit/model';
import openAI from '@genkit-ai/compat-oai';

const modelInfo: ModelInfo = {
  label: 'Claude - Claude 3.7 Sonnet',
  supports: {
    multiturn: true,
    tools: true,
    media: false,
    systemRole: true,
    output: ['json', 'text'],
  },
};
const schema = GenerationCommonConfigSchema.extend({});

const ai = genkit({
  plugins: [
    openAI({
      apiKey: process.env.ANTHROPIC_API_KEY,
      baseURL: 'https://api.anthropic.com/v1/',
      models: [
        { name: 'claude-3-7-sonnet', info: modelInfo, configSchema: schema },
      ],
    }),
  ],
});

export const customModelFlow = ai.defineFlow(
  {
    name: 'customModelFlow',
    inputSchema: z.string(),
    outputSchema: z.string(),
  },
  async (subject) => {
    const llmResponse = await ai.generate({
      prompt: `tell me a joke about ${subject}`,
      model: 'openai/claude-3-7-sonnet',
      config: {
        version: 'claude-3-7-sonnet-20250219',
      },
    });
    return llmResponse.text;
  }
);

Contributing

Want to contribute to the project? That's awesome! Head over to our Contribution Guidelines.

Need help?

• Genkit Discord
• Github Issues

License

This project is licensed under the Apache 2.0 License.