Qwen3-Max-Thinking Benchmark Details

Qwen3-Max-Thinking currently shows benchmark results led by C-Eval (1 / 10, score 93.70), LiveCodeBench (14 / 123, score 85.90), MMLU Pro (22 / 132, score 85.70).

Benchmark Results

Qwen3-Max-Thinking

Benchmark Results

General Knowledge

5 evaluations

Benchmark / mode

Score

Rank/total

C-Eval

93.70

1 / 10

GPQA Diamond

87.40

38 / 187

MMLU Pro

85.70

22 / 132

HLE

49.80

30 / 172

HLE

30.20

88 / 172

Coding and Software Engineer

2 evaluations

Benchmark / mode

Score

Rank/total

LiveCodeBench

85.90

14 / 123

SWE-bench Verified

75.30

37 / 112

Agent Level Benchmark

1 evaluations

Benchmark / mode

Score

Rank/total

τ²-Bench

82.10

11 / 43

Instruction Following

1 evaluations

Benchmark / mode

Score

Rank/total

IF Bench

70.90

12 / 30

Math and Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

IMO-AnswerBench

83.90

11 / 21

Claw-style Agent Evaluation

1 evaluations

Benchmark / mode

Score

Rank/total

Pinch Bench

Thinking EnabledTools

80.30

23 / 37

Compare with other models