归档 / 2026-05-18 / jina-embeddings-v5-omni-small (Jina)

jina-embeddings-v5-omni-small (Jina)

推荐

多模态嵌入模型,支持文本/图像/视频/音频

部署
  • pip pip install sentence-transformers && python -c "from sentence_transformers import SentenceTransformer; model = SentenceTransformer('jinaai/jina-embeddings-v5-omni-small', trust_remote_code=True)"
  • pip pip install transformers && python -c "from transformers import AutoModel; model = AutoModel.from_pretrained('jinaai/jina-embeddings-v5-omni-small', trust_remote_code=True)"
  • vll vllm serve jinaai/jina-embeddings-v5-omni-small --trust-remote-code --hf-overrides '{"task":"retrieval"}'
对位
对位 OpenAI CLIP / Cohere Embed multimodal
适合
多模态检索与RAG / 跨模态零样本分类与聚类
不适合
仅需文本嵌入的轻量场景
规模
1.74B · 32768 · cc-by-nc-4.0
框架
transformers / sentence-transformers / vllm
可信度
下载量 28917,vLLM 0.20.1 已验证,Matryoshka 32-1024 维

评分详情

Q1
今天能接上用吗   3 / 5
Q2
有可信证据吗   3 / 5
Q3
是新东西吗   5 / 5
总分
11
判定
LLM: 推荐  →  规则: 推荐

transformers可用,有详尽示例;下载量2.8万,有arXiv论文;新增多模态embedding能力,与text共享空间。

HuggingFace 原始数据 (抓取于 2026-05-18)

作者
jinaai
任务类型
feature-extraction
推理库
transformers
下载
28,917
点赞
49
许可证
cc-by-nc-4.0
标签
transformers, safetensors, jina_embeddings_v5_omni, feature-extraction, embedding, qwen3, jina-embeddings-v5, sentence-transformers, multimodal, vision, audio, vllm, video, image-feature-extraction, audio-feature-extraction, video-feature-extraction, sentence-similarity, custom_code, multilingual, arxiv:2605.08384, license:cc-by-nc-4.0, region:eu

探索

源链接 ↗