PrimeLLM JavaScript SDK

Official JavaScript/TypeScript SDK for the PrimeLLM unified AI API.

PrimeLLM lets you access multiple AI models (GPT-5.1, Claude, Gemini) through a single, simple API.

Installation

npm install primellm

Quick Start

import PrimeLLM from "primellm";

// Create a client with your API key
const client = new PrimeLLM({
  apiKey: "primellm_XXX", // Get from https://primellm.in/dashboard
});

// Send a chat message
const response = await client.chat({
  model: "gpt-5.1",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "What is TypeScript?" },
  ],
});

// Access the response
console.log(response.choices[0].message.content);
console.log("Tokens used:", response.usage.total_tokens);
console.log("Credits left:", response.credits.remaining);

Available Models

| Model | Description | |-------|-------------| | gpt-5.1 | Latest GPT model (default) | | claude-sonnet-4.5 | Claude Sonnet 4.5 | | gemini-3.0 | Gemini 3.0 |

API Reference

Creating a Client

const client = new PrimeLLM({
  apiKey: "primellm_XXX",  // Required
  baseURL: "https://api.primellm.in",  // Optional, this is the default
  timeoutMs: 60000,  // Optional, 60 seconds default
});

client.chat(request)

Send a chat completion request.

const response = await client.chat({
  model: "gpt-5.1",
  messages: [
    { role: "user", content: "Hello!" }
  ],
  temperature: 0.7,  // Optional
  max_tokens: 1000,  // Optional
});

Response:

{
  id: "chatcmpl_xxx",
  model: "gpt-5.1",
  choices: [{
    index: 0,
    message: { role: "assistant", content: "..." },
    finish_reason: "stop"
  }],
  usage: {
    prompt_tokens: 10,
    completion_tokens: 20,
    total_tokens: 30
  },
  credits: {
    remaining: 149.99,
    cost: 0.00006
  }
}

TypeScript Support

This SDK is written in TypeScript and includes full type definitions.

import PrimeLLM, { ChatRequest, ChatResponse } from "primellm";

const client = new PrimeLLM({ apiKey: "..." });

const request: ChatRequest = {
  model: "gpt-5.1",
  messages: [{ role: "user", content: "Hello!" }],
};

const response: ChatResponse = await client.chat(request);

Error Handling

try {
  const response = await client.chat({
    model: "gpt-5.1",
    messages: [{ role: "user", content: "Hello!" }],
  });
} catch (error) {
  console.error("API Error:", error.message);
}

License

MIT