Alibaba Open weights: Qwen/Qwen3-VL-8B-InstructAuto-detected

Qwen3 Vl 8B

Name: Qwen3 Vl 8B benchmark results
Creator: BenchLeader

Best configuration ranks #440 of 610 on the BenchLeader Index at 43.4 ±1.5 (thinking reasoning effort). Released 15 Oct 2025.

Blended price: $0.660/M; $0.180 in · $2.10 out
Output speed: 111 tok/s
First answer: 20 s; first token 2.46 s
Context: 256k

How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.

Overall index43
Agents & tools37
Knowledge39
Instruction following42
Multimodal41
Long context42
Composite39

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
thinkingbest	43.4	#440	111 tok/s	20 s	$0.0007	Agents & tools 37 · Composite 39 · Instruction following 42 · Knowledge 39 · Long context 42 · Multimodal 41
default	40.4	#508	117 tok/s	2.46 s	$0.0003	Agents & tools 36 · Composite 38 · Instruction following 36 · Knowledge 40 · Long context 33 · Multimodal 31

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmark	thinking	default	Source
GPQA Diamond (AA)not in index	57.9%#369	42.7%#440	Artificial Analysis
Humanity's Last Exam (AA)not in index	3.9%#474	2.7%#520	Artificial Analysis

Agents & tools

Benchmark	thinking	default	Source
Terminal-Bench Hard	3.8%#283	2.3%#304	Artificial Analysis
τ²-Bench Telecom (AA)not in index	22.5%#297	29.2%#256	Artificial Analysis

Knowledge

Benchmark	thinking	default	Source
AA-Omniscience	-52.3#334	-51.9#332	Artificial Analysis

Instruction following

Benchmark	thinking	default	Source
IFBench	39.9%#252	32.3%#327	Artificial Analysis

Multimodal

Benchmark	thinking	default	Source
MMMU-Pro	56.6%#189	47.3%#215	Artificial Analysis

Long context

Benchmark	thinking	default	Source
AA-LCR	33.3%#309	16.7%#375	Artificial Analysis

Composite

Benchmark	thinking	default	Source
AA Intelligence Index	8.2#367	7.3#407	Artificial Analysis

Where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

Provider	Speed	First token	Input $/M	Output $/M	Context	Quantisation
Alibaba Cloud Int.	47 tok/s	0.53 s	$0.117	$0.455	131k	–
Parasail	15 tok/s	1.85 s	$0.250	$0.750	262k	bf16

Price history

Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.

input outputnow $0.117 in · $0.455 out

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.

Workload	Tokens in / out	Cost	With caching	Time
Chat reply	400 / 300	$0.0007	–	22.9 s
Summarise a 30-page report	12,000 / 600	$0.0034	–	25.6 s
Code edit	6,000 / 1,500	$0.0042	–	33.7 s
Agentic coding session	60,000 / 4,000	$0.019	–	56.2 s
Structured extraction	2,000 / 200	$0.0008	–	22.0 s