Back to Explore
Qwen: Qwen3 VL 30B A3B Instruct
Alibaba·Released 2025-10-06
Open Source30B262K ctxApache 2.0Multimodal
About
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...
Related Models
Pricing
Input
$0.13
per 1M tokens
Output
$0.52
per 1M tokens
Blended
$0.23
per 1M tokens
Cheaper than 67% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.23
Monthly
$6.83
vs. Similar Models
Hermes 4 - Llama-3.1 70B (Reasoning)Q:0.0
$0.20-13%
Meta: Llama 4 ScoutQ:0.0
$0.15-34%
Qwen3 14B (Reasoning)Q:+0.1
$0.73+221%
Claude 3.5 Sonnet (Oct '24)Q:-0.1
$6.00+2537%
Performance
120
tokens/sec
Faster than 62% of models
1.04
seconds
Faster than 54% of models
1.04
seconds
Faster than 68% of models
Market Median
94 tok/s
27% faster
Median TTFT
1.11s
6% faster
Throughput/Dollar
525
tok/s per $/1M
Speed Comparison
Grok 4.3 (Non-reasoning)
119 tok/s-0%
Gemini 3 Pro Preview (low)
120 tok/s+0%
Z.ai: GLM 4.7
119 tok/s-1%
Context Window
262K
tokens
Larger than 62% of models
Max Output
33K
tokens
13% of context
Benchmarks
MMLU-Pro
76.4%
GPQA Diamond
69.5%
HLE
6.4%
LiveCodeBench
47.6%
SciCode
30.8%
TerminalBench Hard
6.1%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
72.3%
IFBench
33.1%
Long Context Recall
23.7%
Tau2
19.0%
Market AverageTop Score
Open Source
apache-2.030BGGUF / GPTQ / AWQ
Downloads
586.4K
Likes
581
VRAM (FP16)
24-48 GB
GPU
A6000 / M3 Ultra