Qwen: Qwen3 Max Thinking

Text input Text output
Author's Description

Qwen3-Max-Thinking is the flagship reasoning model in the Qwen3 series, designed for high-stakes cognitive tasks that require deep, multi-step reasoning. By significantly scaling model capacity and reinforcement learning compute, it delivers major gains in factual accuracy, complex reasoning, instruction following, alignment with human preferences, and agentic behavior.

Key Specifications
Cost
$$$$$
Context
262K
Released
Feb 09, 2026
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Response Format Presence Penalty Tools Reasoning Top P Max Tokens Include Reasoning Structured Outputs Seed Tool Choice Temperature
Features

This model supports the following features:

Reasoning Tools Structured Outputs Response Format
Performance Summary

Qwen3 Max Thinking, the flagship reasoning model from Qwen, demonstrates a strong profile for high-stakes cognitive tasks. It exhibits moderate speed performance, ranking in the 39th percentile, and offers moderate pricing, placing it in the 24th percentile. A standout feature is its exceptional reliability, achieving a 100% success rate across all benchmarks, indicating minimal technical failures. The model excels in specific areas, achieving perfect accuracy in Hallucinations (100%) and Ethics (100%), showcasing its ability to acknowledge uncertainty and adhere to ethical principles. It also demonstrates impressive capabilities in Coding (97.0% accuracy, 98th percentile) and Mathematics (96.0% accuracy, 95th percentile), positioning it among the top performers in these technical domains. Reasoning is another strength, with 88.0% accuracy (74th percentile). While its Instruction Following (70.0% accuracy, 77th percentile) and Email Classification (98.0% accuracy, 60th percentile) are solid, its General Knowledge (84.5% accuracy, 23rd percentile) is a relative weakness compared to its other high-performing categories. Overall, Qwen3 Max Thinking is a highly reliable model with exceptional reasoning and technical problem-solving capabilities, making it suitable for demanding applications.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $1.2
Completion $6

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Alibaba
Alibaba | qwen/qwen3-max-thinking-20260123 262K $1.2 / 1M tokens $6 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by qwen