LetTokenBurnBenchHub

Model

Qwen3-VL-4B

Alibaba · open weights

qwenlm.github.io

Coding

Benchmark	Score	Metric	Harness	Source	Confidence	Link
VISTA	3.2	score	N/A	Independent	medium	arxiv.org

Multimodal

Benchmark	Score	Metric	Harness	Source	Confidence	Link
BrainTRACE	32.7	accuracy	N/A	Independent	medium	arxiv.org
FloodVL-Pix	0.3	clair	N/A	Independent	medium	arxiv.org
InstanceBench	54.0	[email protected]	N/A	Independent	medium	arxiv.org
InteractionBench	26.2	score	N/A	Independent	medium	arxiv.org
Meta-Video	55.3	accuracy	N/A	Independent	medium	arxiv.org
OmniVCBench	30.5	accuracy	N/A	Independent	medium	arxiv.org
P^3IBench	28.6	overall_average_prs-b	N/A	Independent	medium	arxiv.org
SpaceCast-Bench	34.0	accuracy	N/A	Independent	medium	arxiv.org
SpatialChain	78.4	accuracy	N/A	Independent	medium	arxiv.org
UkrMeme	0.3	hf1	N/A	Independent	medium	arxiv.org
VisualWorldBench	33.7	accuracy	N/A	Independent	medium	arxiv.org

Reproduced by the community