DataLearner logo
QW

Qwen3-235B-A22B-2507

Chat modelQwen3

Qwen3-235B-A22B-Instruct-2507

Release date: 2025-07-21Updated: 2025-07-27Views: 1,585
Parameters
235B
Context length
256K
Multilingual
Supported
Reasoning ability
3/5

Qwen3-235B-A22B-Instruct-2507 is an AI model published by Alibaba, released on 2025-07-21, for Chat model, with 235B parameters, and 256K context length, requiring about 470.77 GB storage, with a 91.00 score on AIME2025.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Qwen3-235B-A22B-2507

Model basics

Reasoning traces
No data
Thinking modes
Standard Mode (Default)Thinking Mode
Context length
256K tokens
Max output length
32K tokens
Model type
Chat model
Modality (in / out)
Text → Text
Release date
2025-07-21
Model file size
470.77 GB
MoE architecture
Yes
Total params / Active params
235B / 22B
Knowledge cutoff
No data
Qwen3-235B-A22B-2507

Open source & experience

Code license
Weights license
Apache 2.0- Commercial use permitted
Live demo
Qwen3-235B-A22B-2507

Official resources

Paper
DataLearnerAI blog
Qwen3-235B-A22B-2507

API details

API speed
3/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Standard
TypeConditionInputOutput
Text-¥0.0020/ 1K tokens¥0.0080/ 1K tokens

“—” means the modality is not billed in that direction, or the vendor has not published a price for it.

Qwen3-235B-A22B-2507

Benchmark Results

Qwen3-235B-A22B-2507 currently shows benchmark results led by SimpleQA (10 / 47, score 54.30), LiveCodeBench (58 / 250, score 78.80), AIME2025 (50 / 215, score 91). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

5 evaluations
Benchmark / mode
Score
Rank/total
MMLU Pro
Standard Mode
83
47 / 176
LiveBench
Standard Mode
48.84
97 / 117
HLE
Thinking Mode
15.90
326 / 563
ARC-AGI-1
Standard Mode
11
139 / 147
ARC-AGI-2
Standard Mode
1.30
124 / 136

General Evaluation

2 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
77.50
260 / 462
GPQA Diamond
Thinking Mode
79
245 / 462

Common Sense

1 evaluations
Benchmark / mode
Score
Rank/total
SimpleQA
Standard Mode
54.30
10 / 47

Coding and Software Engineer

3 evaluations
Benchmark / mode
Score
Rank/total
LiveCodeBench
Standard Mode
51.80
160 / 250
LiveCodeBench
Thinking Mode
78.80
58 / 250
SciCode
Thinking Mode
41.40
99 / 130

Math and Reasoning

2 evaluations
Benchmark / mode
Score
Rank/total
AIME2025
Standard Mode
70.30
119 / 215
AIME2025
Thinking Mode
91
50 / 215

Agent Level Benchmark

3 evaluations
Benchmark / mode
Score
Rank/total
τ²-Bench - Telecom
Thinking ModeTools
53.20
166 / 264
Terminal Bench Hard
Thinking ModeTools
13.60
173 / 244
τ³-Banking
Thinking ModeTools
7.80
140 / 164

Instruction Following

1 evaluations
Benchmark / mode
Score
Rank/total
IF Bench
Thinking Mode
51.20
151 / 282

AI Agent - Tool Usage

1 evaluations
Benchmark / mode
Score
Rank/total
Terminal-Bench 2.1
Thinking ModeTools
12
170 / 192

Multimodal Understanding

1 evaluations
Benchmark / mode
Score
Rank/total
GDP.pdf
Thinking Mode
4.80
97 / 118
Qwen3-235B-A22B-2507

Model variants & downloads

Variant nameVersion typeQuantizationModel sizeHuggingFace link
Qwen3-235B-A22B-Instruct-2507-FP8ℹ️InstructFP8236.45 GBDownload link
Qwen3-235B-A22B-Instruct-2507ℹ️InstructBF16470.77 GBDownload link
Qwen3-235B-A22B-2507

Publisher

Qwen3-235B-A22B-Instruct-2507

Model Overview

Qwen3-235B-A22B-Instruct-2507 is an AI model published by Alibaba, released on 2025-07-21, for Chat model, with 235B parameters, and 256K context length, requiring about 470.77 GB storage, with a 91.00 score on AIME2025.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code