Qwen3.5 9B

Reasoning

Creator

Alibaba

Released

2026-03-02

Intelligence

32.4

Artificial Analysis Index

Coding

25.3

Artificial Analysis Index

In $/1M

$0.10

input tokens

Out $/1M

$0.15

output tokens

Blended $/1M

$0.11

3:1 blended

Speed

84

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
81
Humanity’s Last Exam
13
MMLU-Pro
SciCode
28
LiveCodeBench
MATH-500
AIME 2025
τ²-Bench (agentic)
87
Terminal-Bench Hard
24
IFBench
67
Via Artificial Analysis