BLINK
BLINK: Multimodal Large Language Models Can See but Not Perceive. A benchmark for multimodal language models focusing on core visual perception abilities. Reformats 14 classic computer vision tasks into 3,807 multiple-choice questions paired with single or multiple images and visual prompting. Tasks include relative depth estimation, visual correspondence, forensics detection, multi-view reasoning, counting, object localization, and spatial reasoning that humans can solve 'within a blink'.
Progress Over Time
Interactive timeline showing model performance evolution on BLINK
State-of-the-art frontier
Open
Proprietary
BLINK Leaderboard
11 models
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 |  Alibaba Cloud / Qwen Team | 236B | 262K | $0.30 / $1.50 | ||
| 2 | Alibaba Cloud / Qwen Team | 9B | 262K | $0.08 / $0.50 | ||
| 3 | Alibaba Cloud / Qwen Team | 9B | 262K | $0.18 / $2.09 | ||
| 4 | Alibaba Cloud / Qwen Team | 33B | — | — | ||
| 5 | Alibaba Cloud / Qwen Team | 31B | 262K | $0.20 / $0.70 | ||
| 6 | Alibaba Cloud / Qwen Team | 33B | — | — | ||
| 7 | Alibaba Cloud / Qwen Team | 236B | 262K | $0.45 / $3.49 | ||
| 8 | Alibaba Cloud / Qwen Team | 4B | 262K | $0.10 / $0.60 | ||
| 9 | Alibaba Cloud / Qwen Team | 31B | 262K | $0.20 / $1.00 | ||
| 10 | Alibaba Cloud / Qwen Team | 4B | 262K | $0.10 / $1.00 | ||
| 11 |  Microsoft | 6B | 128K | $0.05 / $0.10 |
Notice missing or incorrect data?
FAQ
Common questions about BLINK
BLINK: Multimodal Large Language Models Can See but Not Perceive. A benchmark for multimodal language models focusing on core visual perception abilities. Reformats 14 classic computer vision tasks into 3,807 multiple-choice questions paired with single or multiple images and visual prompting. Tasks include relative depth estimation, visual correspondence, forensics detection, multi-view reasoning, counting, object localization, and spatial reasoning that humans can solve 'within a blink'.
The BLINK paper is available at https://arxiv.org/abs/2404.12390. This paper provides detailed information about the benchmark methodology, dataset creation, and evaluation criteria.
The BLINK leaderboard ranks 11 AI models based on their performance on this benchmark. Currently, Qwen3 VL 235B A22B Instruct by Alibaba Cloud / Qwen Team leads with a score of 0.707. The average score across all models is 0.668.
The highest BLINK score is 0.707, achieved by Qwen3 VL 235B A22B Instruct from Alibaba Cloud / Qwen Team.
11 models have been evaluated on the BLINK benchmark, with 0 verified results and 11 self-reported results.
BLINK is categorized under 3d, multimodal, reasoning, spatial reasoning, and vision. The benchmark evaluates multimodal models.