Qwen3 235B A22B 2507 Instruct

Creator

Alibaba

Released

2025-07-21

Intelligence

18.2

Artificial Analysis Index

Coding

Artificial Analysis Index

In $/1M

$0.70

input tokens

Out $/1M

$2.80

output tokens

Blended $/1M

$1.23

3:1 blended

Speed

0

tokens / sec

Profile

License, openness, and modality — the governance layer.

License

Apache 2.0

commercial OK

Weights

Open

downloadable

Modalities

Text

Parameters

235B

total

Qwen3-235B-A22B-Instruct-2507 is the updated instruct version of Qwen3-235B-A22B featuring significant improvements in general capabilities including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. It provides substantial gains in long-tail knowledge coverage across multiple languages and markedly better alignment with user preferences in subjective and open-ended tasks.

Capability profile

Category strength across 20 domains, via LLM Stats.

legal
80biology
80finance
80writing
80language
80creativity
80healthcare
80structured output
80instruction following
80math
70agents
70general
70
physics
70chemistry
70reasoning
70communication
70code
60economics
60tool calling
60vision
50factuality
50multimodal
50spatial reasoning
50
Via LLM Stats

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
75
Humanity’s Last Exam
11
MMLU-Pro
83
SciCode
36
LiveCodeBench
52
MATH-500
98
AIME 2025
72
τ²-Bench (agentic)
33
Terminal-Bench Hard
15
IFBench
46
Via Artificial Analysis
© 2026 NYSGPT2525 LLC