Qwen: Qwen3 VL 30B A3B Instruct

Qwen: Qwen3 VL 30B A3B Instruct

qwen · Released Oct 6, 2025 Efficient
46.2
Our Score

Performance Profile

Intelligence3.6Technical2.6Value7.8Content4
Intelligence 3.6/10
Technical 2.6/10
Content 4/10
Value 7.8/10

Qwen3-VL-30B-A3B-Instruct is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks. It excels in perception..

$0.13 / 1M
Input Price
$0.52 / 1M
Output Price
131,072 tokens
Context Window
32,768 tokens
Max Output
30B Parameters

Capabilities

Tool Use Function Calling Vision

Architecture

ModalityText + Image → Text
TokenizerQwen3
Parameters30B

Performance Indices

Source: Artificial Analysis

16.1 Intelligence Index
14.3 Coding Index
12.6 Agentic Index
72.3 Math Index

Benchmark Scores

Intelligence

GPQA Diamond 69.5% Graduate-level scientific reasoning
HLE 6.4% Humanity's Last Exam
MMLU Pro 76.4% Multi-task language understanding
AIME 2025 72.3% Competition mathematics (2025)
SciCode 30.8% Scientific computing

Technical

LiveCodeBench 47.6% Live coding evaluation
TerminalBench Hard 6.1% Agentic terminal tasks
τ²-Bench 19% Conversational agent benchmark

Content

IFBench 33.1% Instruction following
LCR 23.7% Long-context reasoning

Benchmark data from Artificial Analysis and Hugging Face

How does Qwen: Qwen3 VL 30B A3B Instruct stack up?

Compare side-by-side with other efficient models.

Compare Models

Model Information

OpenRouter ID qwen/qwen3-vl-30b-a3b-instruct
Providerqwen
Release Date October 6, 2025
Context Length131,072 tokens
Max Completion32,768 tokens
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.13 $0.000130
Output $0.52 $0.000520

Live Performance

Live endpoint metrics — refreshed every 30 minutes.

83.1%
Avg Uptime
330ms
Best Latency (TTFT)
56 tok/s
Best Throughput
5/7
Active Endpoints
Available via: Alibaba, DeepInfra, AtlasCloud, Novita, Phala, Venice, SiliconFlow