Skip to main content
Back to Explore

Qwen3 VL 235B A22B (Reasoning)

Alibaba·Released 2025-09-23
Open Source

Pricing

Input

$0.84

per 1M tokens

Output

$6.17

per 1M tokens

Blended

$2.17

per 1M tokens

Cheaper than 24% of models. Median price is $0.54/1M tokens.

Cost Calculator

Tokens per day1M
100K100M

Daily

$2.17

Monthly

$65.22

vs. Similar Models

DeepSeek V3.1 (Reasoning)Q:+0.1
$0.86-60%
Mistral Small 4 (Reasoning)Q:+0.2
$0.26-88%
GPT-5.1 (Non-reasoning)Q:-0.2
$3.44+58%
Nova 2.0 Omni (medium)Q:+0.3
$0.85-61%

Performance

58

tokens/sec

Faster than 26% of models

1.34

seconds

Faster than 39% of models

35.95

seconds

Faster than 14% of models

Market Median

94 tok/s

39% slower

Median TTFT

1.11s

20% slower

Throughput/Dollar

27

tok/s per $/1M

Speed Comparison

Grok Build 0.1 0616
58 tok/s+1%
Anthropic: Claude Opus 4.8
58 tok/s+1%
Qwen: Qwen3 Max
59 tok/s+1%

Benchmarks

MMLU-Pro
83.6%
GPQA Diamond
77.2%
HLE
10.1%
LiveCodeBench
64.6%
SciCode
39.9%
TerminalBench Hard
11.4%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
88.3%
IFBench
56.5%
Long Context Recall
58.7%
Tau2
54.1%
Market AverageTop Score

Open Source

Quick Compare

Similar Models

Compare all 7 models