Benchmark

SimpleVQA

SimpleVQA is a visual question answering benchmark focused on simple queries.

Modality
multimodal
Categories
multimodal, image_to_text, general, vision
Openness
unknown
Source
llm_stats
Reported scores
14

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
GLM-5V-TurboZ.ai0.7822026-04-02evidence
Kimi K2.5Moonshot AI0.7122026-01-27evidence
LFM2.5-VL-3BLiquid AI0.3542026-08-12evidence
Muse SparkMeta0.7132026-04-08evidence
Qwen3 VL 235B A22B ThinkingQwen0.6132025-09-22evidence
Qwen3.5-122B-A10BQwen0.6172026-02-24evidence
Qwen3.5-27BQwen0.562026-02-24evidence
Qwen3.5-35B-A3BQwen0.5832026-02-24evidence
Qwen3.6 PlusQwen0.6732026-04-02evidence
Qwen3.6-27BQwen0.5612026-04-21evidence
Qwen3.6-35B-A3BQwen0.5892026-04-16evidence
Qwen3.7-PlusQwen0.8172026-05-31evidence
Seed 2.1 ProByteDance0.7452026-06-24evidence
Seed 2.1 TurboByteDance0.7112026-06-24evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.