Qwen2.5-Omni-7B
Alibaba Cloud / Qwen TeamQwenOpen WeightApache 2.0 · Commercial OK
描述
Qwen2.5-Omni is the flagship end-to-end multimodal model in the Qwen series. It processes diverse inputs including text, images, audio, and video, delivering real-time streaming responses through text generation and natural speech synthesis using a novel Thinker-Talker architecture.
发布日期
2025-03-27
参数规模
7.0B
上下文长度
—
支持模态
—
能力雷达图
50
general
80
coding
60
reasoning
26
science估算
0
agents
90
multimodal
Science 在缺少专门科学评测时使用推理能力代理估算。
排行榜排名
| 领域 | #排名 | 分数 | 来源 |
|---|---|---|---|
| 多模态榜 | 52 | 74.0 | LS |
基准测试分数 (LLM Stats)
Audio
VocalSound
93.9%自报
GiantSteps Tempo
88.0%自报
MMAU Music
69.2%自报
MMAU Sound
67.9%自报
MMAU
65.6%自报
MMAU Speech
59.8%自报
OmniBench Music
52.8%自报
CoVoST2 en-zh
0.41 / 100自报
MusicCaps
32.8%自报
Common Voice 15
0.08 / 100自报
Biology
GPQA
30.8%自报
Code
HumanEval
78.7%自报
Communication
VoiceBench Avg
74.1%自报
MM-MT-Bench
0.06 / 100自报
Creativity
Meld
57.0%自报
Finance
MMLU-Pro
47.0%自报
General
MBPP
0.73 / 100自报
MMLU-Redux
71.0%自报
MultiPL-E
65.8%自报
MMStar
64.0%自报
MME-RealWorld
61.6%自报
MMMU
59.2%自报
MMMU-Pro
36.6%自报
LiveBench
29.6%自报
NMOS
0.05 / 100自报
Grounding
PointGrounding
66.5%自报
Healthcare
CRPErelation
76.5%自报
Image To Text
DocVQA
95.2%自报
TextVQA
84.4%自报
OCRBench_V2
57.8%自报
Language
FLEURS
0.04 / 100自报
Long Context
EgoSchema
68.6%自报
Math
GSM8k
88.7%自报
MATH
71.5%自报
MathVista
67.9%自报
MathVision
25.0%自报
Multimodal
ChartQA
85.3%自报
AI2D
83.2%自报
MMBench-V1.1
81.8%自报
VideoMME w sub.
72.4%自报
MVBench
70.3%自报
MuirBench
59.2%自报
OmniBench
56.1%自报
Spatial Reasoning
RealWorldQA
70.3%自报
Vision
ODinW
42.4%自报
AA 评测指数
暂无 AA 评测数据
LLM Stats 分类评分
Image To Text90
Code80
Spatial Reasoning70
Video70
Vision70
Long Context70
Math60
Multimodal60
Reasoning60
Finance50
General50
Healthcare50
Language50
Legal50
Biology30
Chemistry30
Physics30
Communication10
Speech To Text0
定价
暂无定价数据
速度
暂无速度数据
可用提供商
(LS 内部计价单位)暂无提供商数据