---
last_updated: 2026-11-20
source_url: https://opper.ai/alibaba/qwen3-max-thinking
title: "Qwen3 Max Thinking by Alibaba"
description: "Qwen3 Max Thinking by Alibaba, a proprietary reasoning model with 256K context and adaptive tool use for complex problem solving."
---

# Qwen3 Max Thinking

By [Alibaba](https://opper.ai/alibaba).

Released 23 Jan 2026, 1 days after [Qwen3-TTS](https://opper.ai/alibaba/qwen3-tts) · [Announcement](https://www.alibabacloud.com/help/en/model-studio/newly-released-models) · [All Alibaba releases](https://opper.ai/model-releases/alibaba)

Qwen3 Max Thinking is Alibaba's trillion-parameter reasoning model that combines step-by-step thinking with adaptive tool use, selecting and invoking tools such as code interpreters during inference. The model generates extended reasoning chains before its final answer, making intermediate reasoning visible.

Reported to reach perfect scores on challenging math benchmarks including AIME 2025 and HMMT when paired with tool use and scaled test-time compute, this model targets complex mathematics, logic, code synthesis, and agentic planning. Its thinking mode surfaces explicit problem-solving steps, which helps with interpretability and multi-step verification.

It is best suited to developers building reasoning-intensive agents, educational AI systems, and applications where understanding the model's thought process matters as much as the final answer. With a 256K context window, it can reason over long inputs while keeping its working steps available for inspection.

## Available routes

Qwen3 Max Thinking runs on 1 route through the Opper gateway. Prices are per 1M tokens. Its routes and prices, live from the catalogue, are public JSON with no API key needed at https://api.opper.ai/v3/models?q=deepinfra%2FQwen%2FQwen3-Max-Thinking.

| Provider | Model id | Region | Zero data retention | Training | Input | Output | Cache read |
| --- | --- | --- | --- | --- | --- | --- | --- |
| [DeepInfra](https://opper.ai/provider/deepinfra) | `deepinfra/Qwen/Qwen3-Max-Thinking` | US (United States) | Yes | No | $1.20 | $6.00 | $0.24 |

## Key info

- Model id: `deepinfra/Qwen/Qwen3-Max-Thinking`
- Input: Text
- Output: Text
- Features: Tools
- Context window: 256K
- Max output: 256K
- Input price: $1.20 per 1M tokens ($0.24 cache read)
- Output price: $6.00 per 1M tokens
- Released: 23 Jan 2026
- US residency available
- Zero data retention on pay-as-you-go
- No training by default

## Data handling per route

Each route hosting Qwen3 Max Thinking has its own privacy posture, residency, and GDPR terms. Postures are maintained by Opper with a last-verification timestamp.

### DeepInfra — United States

Zero data retention: nothing is logged, held for abuse monitoring or used for training. No training on customer data. US; SCCs.

- Zero data retention: Yes. Nothing is logged, held for abuse monitoring or used for training.
- Training: No training on customer data.
- Logging: None
- Abuse monitoring: No classifier
- Transfer mechanism: SCCs

Verified by Opper on 2026-10-19.

## Benchmarks

Independent benchmark scores via [Artificial Analysis](https://artificialanalysis.ai) ([view on AA](https://artificialanalysis.ai/models/qwen3-max-thinking)). Indices and domain scores are out of 100, Omniscience runs from −100 to 100, Briefcase is an Elo rating, and the other evaluation scores are the share of tasks solved.

- Global rank: #169 of 520 LLMs on the [LLM leaderboard](https://opper.ai/llm-leaderboard)
- Tier: Efficient
- Intelligence Index (previous version): 21.3

### Agentic & knowledge work

- Terminal-Bench Hard: 24%
- τ²-Bench Telecom: 84%

### Science & reasoning

- Humanity's Last Exam: 28%
- GPQA Diamond: 86%
- Long-context reasoning: 74%

### Math & instruction following

- IFBench: 71%

## Get started

Call Qwen3 Max Thinking through the Opper gateway with one API key. Opper is drop-in compatible with the OpenAI, Anthropic and Google AI SDKs.

Qwen3 Max Thinking is a premium model on Opper. Sign up needs no credit card: you get an API key straight away and the free models work in the playground and the API. Add a card to use premium models, pay-as-you-go with no minimum.

### Set it up with your agent

Paste this into a coding agent like Claude Code, Cursor or Codex and it will wire up Opper for you:

```text
Use curl to download, read and follow: https://skills.opper.ai Call Qwen3 Max Thinking
```

### Or call it directly

OpenAI:

```javascript
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.OPPER_API_KEY,
  baseURL: "https://api.opper.ai/v3/compat",
});

const completion = await client.chat.completions.create({
  model: "deepinfra/Qwen/Qwen3-Max-Thinking",
  messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);
```

Anthropic:

```javascript
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: process.env.OPPER_API_KEY,
  baseURL: "https://api.opper.ai/v3/compat",
});

const message = await client.messages.create({
  model: "deepinfra/Qwen/Qwen3-Max-Thinking",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello" }],
});
console.log(message.content);
```

Gemini:

```javascript
import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({
  apiKey: process.env.OPPER_API_KEY,
  httpOptions: { baseUrl: "https://api.opper.ai/v3/compat" },
});

const response = await ai.models.generateContent({
  model: "deepinfra/Qwen/Qwen3-Max-Thinking",
  contents: "Hello",
});
console.log(response.text);
```

cURL:

```bash
curl https://api.opper.ai/v3/compat/chat/completions \
  -H "Authorization: Bearer $OPPER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepinfra/Qwen/Qwen3-Max-Thinking",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'
```

## Use Qwen3 Max Thinking in your apps

Pick your tool to see how to run it on Qwen3 Max Thinking through Opper.

- [Claude Code](https://opper.ai/apps/claude-code/qwen3-max-thinking)
- [Claude Cowork](https://opper.ai/apps/claude-cowork/qwen3-max-thinking)
- [Codex](https://opper.ai/apps/codex/qwen3-max-thinking)
- [DeepSeek Harness](https://opper.ai/apps/deepseek-harness/qwen3-max-thinking)
- [GitHub Copilot CLI](https://opper.ai/apps/github-copilot-cli/qwen3-max-thinking)
- [Aider](https://opper.ai/apps/aider/qwen3-max-thinking)
- [Goose](https://opper.ai/apps/goose/qwen3-max-thinking)
- [Droid](https://opper.ai/apps/droid/qwen3-max-thinking)
- [OpenCode](https://opper.ai/apps/opencode/qwen3-max-thinking)
- [Pi](https://opper.ai/apps/pi/qwen3-max-thinking)
- [OpenClaw](https://opper.ai/apps/openclaw/qwen3-max-thinking)
- [Hermes](https://opper.ai/apps/hermes/qwen3-max-thinking)
- [Copilot for Obsidian](https://opper.ai/apps/obsidian-copilot/qwen3-max-thinking)
- [OmniRoute](https://opper.ai/apps/omniroute/qwen3-max-thinking)
- [Cursor](https://opper.ai/apps/cursor/qwen3-max-thinking)
- [Cline](https://opper.ai/apps/cline/qwen3-max-thinking)
- [Continue](https://opper.ai/apps/continue/qwen3-max-thinking)
- [Kilo Code](https://opper.ai/apps/kilo-code/qwen3-max-thinking)
- [GitHub Copilot](https://opper.ai/apps/github-copilot/qwen3-max-thinking)
- [Zed](https://opper.ai/apps/zed/qwen3-max-thinking)
- [JetBrains AI Assistant](https://opper.ai/apps/jetbrains-ai-assistant/qwen3-max-thinking)
- [Voidleap Code](https://opper.ai/apps/voidleap-code/qwen3-max-thinking)
- [n8n](https://opper.ai/apps/n8n/qwen3-max-thinking)
- [LibreChat](https://opper.ai/apps/librechat/qwen3-max-thinking)
- [Open WebUI](https://opper.ai/apps/open-webui/qwen3-max-thinking)
- [See all apps](https://opper.ai/apps)

## Other models from Alibaba

- [Qwen4-Max](https://opper.ai/alibaba/qwen4-max)
- [Qwen3.8 Max](https://opper.ai/alibaba/qwen3-8-max)
- [Qwen3.8 27B](https://opper.ai/alibaba/qwen3-8-27b)
- [Qwen3.8-2.4T-A95B](https://opper.ai/alibaba/qwen3-8-2-4t-a95b)
- [Qwen3.8 Flash-Next](https://opper.ai/alibaba/qwen3-8-flash-next)
- [Qwen3.7-Max](https://opper.ai/alibaba/qwen3-7-max)
