modelgrep
Q

Qwen: Qwen3 Max Thinking

qwen/qwen3-max-thinking

41st smartest of 180ReasoningToolsJSON
Use via OpenRouter ↗
Intelligence
39.8
41st of 180
Design Elo
Speed
40
202nd fastest
Latency
1.1s
first token
Input price
$0.780
190th cheapest
Context
262K
33K max out

How it compares

Smarter than77%
of all ranked models
Faster than34%
of all ranked models
Cheaper than36%
of all ranked models

Overview

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...

Benchmarks

independent · via OpenRouter
Artificial Analysis73th percentile
Intelligence Index
39.8
Coding Index
30.5
Agentic Index
50.1
GPQA Diamond
86%
Humanity's Last Exam
26%
SciCode
43%
Tau²-Bench (agentic)
84%

Providers & pricing (1)

ProviderIn $/MOut $/MUptime
Alibaba$0.780$3.90

Specifications

Context window262K
Max output33K
Knowledge cutoff
Input modalitiestext
Output modalitiestext
Prompt caching
Cache read price
ModeratedNo

Qwen3 Max Thinking FAQ

How much does Qwen3 Max Thinking cost?

Qwen3 Max Thinking costs $0.780 per million input tokens and $3.90 per million output tokens via OpenRouter, making it 190th cheapest of 298 paid models.

How smart is Qwen3 Max Thinking?

Qwen3 Max Thinking scores 39.8 on the Artificial Analysis Intelligence Index, ranking 41st of 180 benchmarked models, with a GPQA Diamond score of 86%.

How fast is Qwen3 Max Thinking?

Qwen3 Max Thinking generates around 40 tokens per second with 1.1s time-to-first-token (p50), the 202nd fastest tracked model.

What is Qwen3 Max Thinking's context window?

Qwen3 Max Thinking supports a 262K-token context window and can output up to 33K tokens. It accepts text input.

Compare head-to-head