Skip to main content
Back to Explore

Qwen3 VL 8B (Reasoning)

Alibaba·Released 2025-10-14
Open Source
Compare

Related Models

Qwen: Qwen3.7 Plus2026-06-03Qwen: Qwen3.7 Max2026-05-21Qwen: Qwen3.6 27B2026-04-27Qwen: Qwen3.6 35B A3B2026-04-27Qwen: Qwen3.5 Plus 2026-04-202026-04-27Qwen: Qwen3.6 Flash2026-04-27Qwen: Qwen3.6 Max Preview2026-04-27Qwen3.6 27B (Non-reasoning)2026-04-22

Pricing

Input

$0.18

per 1M tokens

Output

$2.10

per 1M tokens

Blended

$0.66

per 1M tokens

Cheaper than 47% of models. Median price is $0.54/1M tokens.

Cost Calculator

Tokens per day1M
100K100M

Daily

$0.66

Monthly

$19.80

vs. Similar Models

Nova 2.0 Omni (Non-reasoning)Q:-0.1
$0.85+29%
Qwen3 32B (Reasoning)Q:-0.1
$0.28-58%
DeepSeek V3 (Dec '24)Q:-0.2
$0.52-21%
Qwen3 235B A22B (Non-reasoning)Q:+0.3
$0.79+19%

Performance

135

tokens/sec

Faster than 67% of models

1.06

seconds

Faster than 53% of models

15.87

seconds

Faster than 30% of models

Market Median

94 tok/s

43% faster

Median TTFT

1.11s

5% faster

Throughput/Dollar

205

tok/s per $/1M

Speed Comparison

Z.ai: GLM 5.2
135 tok/s+0%
Anthropic: Claude 3 Haiku
134 tok/s-1%
OpenAI: GPT-4.1
134 tok/s-1%

Benchmarks

MMLU-Pro
74.9%
GPQA Diamond
57.9%
HLE
3.3%
LiveCodeBench
35.3%
SciCode
21.9%
TerminalBench Hard
3.8%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
30.7%
IFBench
39.9%
Long Context Recall
31.0%
Tau2
22.5%
Market AverageTop Score

Open Source

Quick Compare

Similar Models

EXAONE 4.0 32B (Reasoning)

LG AI Research

Q: 10.6N/A/1M
Coding: +4.2

Nova 2.0 Omni (Non-reasoning)

Amazon

Q: 10.5$0.85/1M
Faster: 62%Pricier: 29%

Gemini 2.0 Flash (experimental)

Google

Q: 10.7N/A/1M

Magistral Small 1

Mistral

Q: 10.7N/A/1M

Qwen3 32B (Reasoning)

Alibaba

Q: 10.5$0.28/1M
Slower: 31%Cheaper: 58%

DeepSeek R1 0528 Qwen3 8B

DeepSeek

Q: 10.4N/A/1M
Compare all 7 models