AiCostCompare

Qwen3 VL 30B A3B Instruct

by Alibaba (Qwen) · qwen/qwen3-vl-30b-a3b-instruct · #69 cheapest of 325 paid models

Prices updated Jul 22, 2026, 2:26 AM UTC · refreshed hourly

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...

Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol

Pricing

Input

$0.13

per 1M tokens

Output

$0.52

per 1M tokens

Blended (3:1)

$0.228

3 input : 1 output

Cache read

per 1M cached tokens

Cache write

per 1M tokens

Cache & batch economics

Effective prices for Qwen3 VL 30B A3B Instruct (no cache-read rate published — hits billed as normal input).

Cache hit rateEffective blended $/1MExample request*vs no cache
0%$0.228$0.0138
50%$0.228$0.0138−0%
90%$0.228$0.0138−0%

*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.

What a request costs

WorkloadInput tokensOutput tokensCost / requestCost / 1K requests
Short chat message500300$0.000221$0.221
Document summary8,0001,000$0.00156$1.56
Codebase question (RAG)30,0002,000$0.00494$4.94
Long-context analysis150,0005,000$0.0221$22.10

Performance

Intelligence Index

10

AA composite quality

Coding Index

Math Index

72.3

Agentic Index

Output speed

126 tok/s

median

Time to first token

0.95s

median

Time to first answer

0.95s

after reasoning tokens

Artificial Analysis benchmarks

BenchmarkScore
MMLU-Pro76.4%
GPQA Diamond69.5%
Humanity's Last Exam6.4%
LiveCodeBench47.6%
SciCode30.8%
AIME 202572.3%
IFBench33.1%
AA-LCR (Long Context Reasoning)23.7%
Terminal-Bench Hard6.1%
τ²-Bench (Telecom)19.0%

Specs

Context window

262K

tokens

Max output

33K

tokens

Input modalities

text, image

Tokenizer

Qwen3