Qwen·Qwen 3 family

Qwen 3.8 Max

MoEproprietaryReleased Aug 2026

Alibaba's flagship next-generation Qwen model, a sparse Mixture-of-Experts system with roughly 2.4 trillion total parameters and a 1M-token context window. Natively multimodal across text, images, video, and documents, with a configurable thinking mode, function calling, and built-in tools. Positioned by Alibaba as second only to Claude Fable 5 among frontier models, with particular strength in long-context agentic coding, document work, and visual reasoning. Previewed at the World AI Conference in Shanghai on 19 July 2026 and made broadly available on 3 August 2026.

Capabilities

ChatText completionFunction callingVisionCode generationReasoning

Modalities

textimagevideocode

Context window

1.0M tokens

Providers

Available

Available from 1 provider

Cheapest

Alibaba Model Studio

$8.00/1M tokens

Alibaba Model Studio

Provider (1)

Sorted by total cost (input + output per 1M tokens). Select a row to view provider details.

Provider	Pricing (per 1M)	Rate limits	Regions	Health	Latency
Alibaba Model Studio	In: $2.00Out: $6.00	300 RPM / 500K TPM	ap-east-1	Healthy	0ms

Quick start

Use this model via Alibaba Model Studio with an OpenAI-compatible SDK.

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.alibaba-model-studio.com/v1",
  apiKey: process.env.ALIBABA_MODEL_STUDIO_API_KEY,
});

const response = await client.chat.completions.create({
  model: "qwen3.8-max",
  messages: [
    { role: "user", content: "Hello!" }
  ],
});

console.log(response.choices[0].message.content);

Using Alibaba Model Studio API · OpenAI-compatible SDK