CortiqaDevelopers

Developer Tooling

Official SDKs & Client Libraries

Cortiqa provides production-ready official client libraries across 6 major languages and runtimes. Each SDK is optimized for high-throughput inference on Cortiqa LPU infrastructure with native token streaming, automatic retries, and typed exceptions.

Official SDK Matrix

Language / RuntimePackage NameInstallationRepository
Pythoncortiqapip install cortiqacortiqa-sdk-python
TypeScript / Node.js@cortiqa/sdknpm install @cortiqa/sdkcortiqa-sdk-typescript
Gocortiqa-sdk-gogo get github.com/cortiqa-ai/cortiqa-sdk-gocortiqa-sdk-go
Java / Androidcortiqa-sdk-javaco.cortiqa:cortiqa-sdk-java:0.1.0cortiqa-sdk-java
C# / .NETCortiqa.Sdkdotnet add package Cortiqa.Sdkcortiqa-sdk-csharp
Swift (Apple Platforms)cortiqa-sdk-swiftSwift Package Manager (0.1.0+)cortiqa-sdk-swift

1. Python SDK

The official Python client supports synchronous and high-concurrency asynchronous execution via AsyncCortiqa:

Terminal
pip install cortiqa
quickstart.py
1import sys
2from cortiqa import Cortiqa
3
4# Automatically reads CORTIQA_API_KEY from environment
5client = Cortiqa()
6
7# Basic Chat Completion
8response = client.chat.completions.create(
9    model="openai/gpt-oss-120b",
10    messages=[
11        {"role": "system", "content": "You are a helpful assistant by Cortiqa."},
12        {"role": "user", "content": "Explain quantum computing in two sentences."}
13    ],
14    temperature=0.7,
15)
16print(response.choices[0].message.content)
17
18# Real-Time Token Streaming
19stream = client.chat.completions.create(
20    model="openai/gpt-oss-120b",
21    messages=[{"role": "user", "content": "Write a short poem about space."}],
22    stream=True,
23)
24for chunk in stream:
25    if chunk.choices and chunk.choices[0].delta.content:
26        sys.stdout.write(chunk.choices[0].delta.content)
27        sys.stdout.flush()

2. TypeScript & Node.js SDK

Works across Node.js (18+), Next.js App Router, Cloudflare Workers, Vercel Edge, and Bun:

Terminal
npm install @cortiqa/sdk
index.ts
1import Cortiqa from "@cortiqa/sdk";
2
3const client = new Cortiqa(); // Reads process.env.CORTIQA_API_KEY
4
5async function main() {
6  // Option A: Messages API
7  const message = await client.messages.create({
8    model: "openai/gpt-oss-120b",
9    max_tokens: 1024,
10    messages: [{ role: "user", content: "Explain quantum computing in two sentences." }],
11  });
12  console.log(message.content);
13
14  // Option B: Real-time token streaming
15  const stream = await client.messages.stream({
16    model: "openai/gpt-oss-120b",
17    messages: [{ role: "user", content: "Write a poem about Mumbai." }],
18  });
19
20  for await (const token of stream.textStream()) {
21    process.stdout.write(token);
22  }
23}
24
25main();

3. Go SDK

Built with pure Go standard library and zero external dependencies:

Terminal
go get github.com/cortiqa-ai/cortiqa-sdk-go
main.go
1package main
2
3import (
4	"context"
5	"fmt"
6	"log"
7
8	"github.com/cortiqa-ai/cortiqa-sdk-go"
9)
10
11func main() {
12	client := cortiqa.NewClient("") // Reads CORTIQA_API_KEY env var
13	ctx := context.Background()
14
15	resp, err := client.Chat.Create(ctx, cortiqa.ChatCompletionRequest{
16		Model: "openai/gpt-oss-120b",
17		Messages: []cortiqa.ChatMessage{
18			{Role: "system", Content: "You are a concise AI assistant by Cortiqa."},
19			{Role: "user", Content: "Explain Goroutines in 2 sentences."},
20		},
21	})
22	if err != nil {
23		log.Fatalf("Error: %v", err)
24	}
25
26	fmt.Println(resp.Content())
27	fmt.Printf("Total tokens: %d\n", resp.Usage.TotalTokens)
28}

4. Java SDK

Available on Maven Central for Java 11+ and Spring Boot:

pom.xml
<dependency>
    <groupId>co.cortiqa</groupId>
    <artifactId>cortiqa-sdk-java</artifactId>
    <version>0.1.0</version>
</dependency>
Main.java
1import co.cortiqa.sdk.CortiqaClient;
2import co.cortiqa.sdk.models.*;
3
4public class Main {
5    public static void main(String[] args) {
6        // Reads from CORTIQA_API_KEY environment variable
7        CortiqaClient client = CortiqaClient.fromEnv();
8
9        ChatCompletionResponse response = client.messages().create(
10            ChatCompletionRequest.builder()
11                .model("openai/gpt-oss-120b")
12                .addMessage(ChatMessage.user("Explain reactive microservices in 2 sentences."))
13                .build()
14        );
15
16        System.out.println(response.getContent());
17    }
18}

5. C# / .NET SDK

Native async/await, IAsyncEnumerable streaming, and ASP.NET Core dependency injection:

Terminal
dotnet add package Cortiqa.Sdk
Program.cs
1using System;
2using Cortiqa.Sdk;
3using Cortiqa.Sdk.Models;
4
5// Automatically reads from CORTIQA_API_KEY
6using var client = CortiqaClient.FromEnvironment();
7
8var response = await client.Messages.CreateAsync(
9    ChatCompletionRequest.Builder()
10        .Model("openai/gpt-oss-120b")
11        .AddMessage(ChatMessage.User("Explain dependency injection in C#."))
12        .Build()
13);
14
15Console.WriteLine(response.Content);

6. Swift SDK (iOS, macOS, visionOS)

Add via Swift Package Manager: https://github.com/cortiqa-ai/cortiqa-sdk-swift.git

ContentView.swift
1import Cortiqa
2
3let client = CortiqaClient.fromEnvironment()
4
5let response = try await client.messages.create(
6    ChatCompletionRequest(
7        model: "openai/gpt-oss-120b",
8        messages: [
9            .user("Explain SwiftUI state management in 2 sentences.")
10        ]
11    )
12)
13
14print(response.content)

OpenAI Client Drop-in Compatibility

If you already have applications built with standard OpenAI client libraries in any language, you can switch to Cortiqa simply by pointing base_url to https://api.cortiqa.co/api/v1:

dropin.py
from openai import OpenAI

client = OpenAI(
    api_key="sk-cortiqa-YOUR_API_KEY",
    base_url="https://api.cortiqa.co/api/v1"
)

response = client.chat.completions.create(
    model="openai/gpt-oss-120b",
    messages=[{"role": "user", "content": "Hello from existing OpenAI client!"}]
)
print(response.choices[0].message.content)
Was this page helpful?
Previous
Quickstart
Next
Models