Jovethra / OpenAI-compatible

Start with the client you already use.

These are starting points for an OpenAI-compatible base URL. Keep the API key on the server, begin with a non-sensitive fixture, and validate the exact version of each tool before production.

Ready-to-adapt templates

One endpoint, familiar primitives.

OpenAI SDK

Use Responses for new integrations and keep the key in an environment variable.

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.JOVETHRA_API_KEY,
  baseURL: "https://api.jovethra.xyz/v1"
});

await client.responses.create({
  model: "gpt-5.6-sol",
  store: false,
  input: "Run a non-sensitive fixture."
});

Download starter →

Vercel AI SDK

Create a custom OpenAI-compatible provider, then supply the public model ID.

import { createOpenAI } from "@ai-sdk/openai";
import { generateText } from "ai";

const jovethra = createOpenAI({
  apiKey: process.env.JOVETHRA_API_KEY,
  baseURL: "https://api.jovethra.xyz/v1"
});

await generateText({
  model: jovethra("gpt-5.6-sol"),
  prompt: "Run a non-sensitive fixture."
});

Download starter →

LangChain

Pass the compatible base URL through the OpenAI chat client configuration.

import { ChatOpenAI } from "@langchain/openai";

const model = new ChatOpenAI({
  model: "gpt-5.6-sol",
  apiKey: process.env.JOVETHRA_API_KEY,
  configuration: {
    baseURL: "https://api.jovethra.xyz/v1"
  }
});

Download starter →

Continue and Cursor

Choose an OpenAI-compatible provider in the tool settings, then use the base URL and a public model ID. Configuration formats change between releases.

Base URL
https://api.jovethra.xyz/v1

Model
gpt-5.6-sol

API key
JOVETHRA_API_KEY

Check the API contract →

Cursor, Continue, LangChain, and Vercel AI SDK evolve independently. These snippets are starting configurations, not a certification of every release or feature. Test tool calls, streaming, and error handling with your own version before production.

For agent workloads

Model selection and hard limits.

Use Sol as the documented baseline. Terra is also available through the API and consumes lower weighted cost units. Input, output, cache, request-rate, and concurrency limits continue to apply independently of the client library.

Read quota behavior →