Qwen3-Omni Technical Report

Benchmark Model Rank Results
code-generation-on-multipl-eQwen3-Omni-Flash-Instruct#2Accuracy: 81.5
text-generation-on-writingbenchQwen3-Omni-Flash-Thinking#2Score (0-10): 8.59
audio-understanding-reasoning-on-mmsuQwen3-Omni-Flash-Thinking#4Accuracy: 71.3
multi-task-language-understanding-on-livebench-2024-11-25Qwen3-Omni-Flash-Thinking#4Score: 70.3
visual-question-answering-vqa-on-ai2d-test-with-maskQwen3-Omni-Flash-Thinking#4Accuracy: 86.8
audio-visual-question-answering-on-worldsenseQwen3-Omni-30B-A3B-Instruct#5Accuracy: 54.0
audio-visual-question-answering-on-dailyomniQwen3-Omni-30B-A3B-Thinking#6Accuracy: 75.8
video-question-answering-on-videoholmesQwen3-Omni-Flash-Thinking#7Accuracy: 57.3
logic-puzzle-solving-on-zebralogicQwen3-Omni-Flash-Instruct#9Accuracy: 76.1
visual-question-answering-vqa-on-hallusionbench-average-accuracyQwen3-Omni-Flash-Thinking#10Accuracy: 63.4
object-counting-on-countbenchQwen3-Omni-Flash-Thinking#13Accuracy: 92.5
chart-question-answering-on-chartqaQwen3-Omni-Flash-Thinking#141:1 Accuracy: 89.3
multi-task-language-understanding-on-mmlu-reduxQwen3-Omni-Flash-Thinking#16Accuracy: 89.7
multimodal-reasoning-on-mmmuQwen3-Omni-Flash-Thinking#17Accuracy: 75.0
multimodal-reasoning-on-mathvisionQwen3-Omni-Flash-Thinking#21Accuracy: 63.8
video-question-answering-on-mlvuQwen3-Omni-Flash-Instruct#21Accuracy: 75.7
vision-language-understanding-on-mmstarQwen3-Omni-Flash-Thinking#22Accuracy: 75.5
zero-shot-video-question-answer-on-video-mme-w-o-subsQwen3-Omni-Flash-Instruct#22Accuracy (%): 71.4
multimodal-reasoning-on-mathvistaQwen3-Omni-Flash-Thinking#24Accuracy: 81.2
multimodal-reasoning-on-mmmu-proQwen3-Omni-Flash-Thinking#28Accuracy: 60.8
video-question-answering-on-lvbenchQwen3-Omni-Flash-Instruct#32Accuracy: 51.1
mathematical-reasoning-on-aime-2025Qwen3-Omni-Flash-Thinking#48Accuracy: 74.0
multiple-choice-question-answering-mcqa-on-gpqa-diamondQwen3-Omni-Flash-Thinking#78Accuracy: 73.1