multimodal benchmark
MTVQA
MTVQA (Multilingual Text-Centric Visual Question Answering) is the first benchmark featuring high-quality human expert annotations across 9 diverse languages, consisting of 6,778 question-answer pairs across 2,116 images. It addresses visual-textual misalignment problems in multilingual text-centric VQA.
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
MTVQA Ranking
Higher score ranks better on this benchmark.
| 01 | AC | 30.9% | 100.0% | 1 | C |
MTVQA Highlights
The leading models and scores on this benchmark.
What is MTVQA?
What MTVQA measures and how its scores work.
MTVQA (Multilingual Text-Centric Visual Question Answering) is the first benchmark featuring high-quality human expert annotations across 9 diverse languages, consisting of 6,778 question-answer pairs across 2,116 images. It addresses visual-textual misalignment problems in multilingual text-centric VQA.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MTVQA
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mtvqa|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MTVQA.
Which model scores highest on MTVQA?
Qwen2-VL-72B-Instruct is currently ranked first with 30.9%.
What does MTVQA measure?
MTVQA (Multilingual Text-Centric Visual Question Answering) is the first benchmark featuring high-quality human expert annotations across 9 diverse languages, consisting of 6,778 question-answer pairs across 2,116 images. It addresses visual-textual misalignment problems in multilingual text-centric VQA.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.