About
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...
Related Models
Pricing
Input
$0.78
per 1M tokens
Output
$3.90
per 1M tokens
Blended
$1.56
per 1M tokens
Cheaper than 28% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$1.56
Monthly
$46.80
vs. Similar Models
Qwen: Qwen3.6 35B A3BQ:-0.1
$0.35-77%
Qwen3.5 397B A17B (Non-reasoning)Q:+0.3
$1.35-13%
MiniMax: MiniMax M2.1Q:-0.3
$0.45-71%
DeepSeek V4 Pro (Non-reasoning)Q:-0.5
$0.54-65%
Performance
44
tokens/sec
Faster than 10% of models
1.45
seconds
Faster than 35% of models
46.83
seconds
Faster than 7% of models
Market Median
94 tok/s
53% slower
Median TTFT
1.11s
30% slower
Throughput/Dollar
28
tok/s per $/1M
Speed Comparison
MiMo-V2-Pro
44 tok/s+1%
QwQ 32B-Preview
43 tok/s-2%
DeepSeek R1 Distill Qwen 32B
43 tok/s-3%
Context Window
262K
tokens
Larger than 62% of models
Max Output
33K
tokens
13% of context
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
86.1%
HLE
26.2%
LiveCodeBenchNot evaluated
SciCode
43.1%
TerminalBench Hard
24.2%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
70.7%
Long Context Recall
66.0%
Tau2
83.6%
Market AverageTop Score