Category: provider
Model Studio Multimodal Embedding
Validation
CODEBLOCK0
Pass criteria: command exits 0 and output/aliyun-qwen-multimodal-embedding/validate.txt is generated.
Output And Evidence
- Save normalized request payloads, selected dimensions, and sample input references under output/aliyun-qwen-multimodal-embedding/. Record the exact model, modality mix, and output vector dimension for reproducibility. Use this skill when the task needs text, image, or video embeddings from Model Studio for retrieval or similarity workflows.
Critical model names
Use one of these exact model strings as needed:
- INLINECODE2 INLINECODE3 INLINECODE4 Selection guidance:
- Prefer qwen3-vl-embedding for the newest multimodal embedding path. Use qwen2.5-vl-embedding when you need compatibility with an older deployed pipeline. Prerequisites
- Set DASHSCOPE_API_KEY in your environment, or add dashscope_api_key to ~/.alibabacloud/credentials. Pair this skill with a vector store such as DashVector, OpenSearch, or Milvus when building retrieval systems. Normalized interface (embedding.multimodal)
Request
- model (string, optional): default INLINECODE11 INLINECODE12 (array, optional) INLINECODE13 (array, optional): public URLs or local paths uploaded by your client layer INLINECODE14 (array, optional): public URLs where supported INLINECODE15 (int, optional): e.g. 2560, 2048, 1536, 1024, 768, 512, 256 for INLINECODE23 Response
- embeddings (array) INLINECODE25 (int) INLINECODE26 (object, optional) Quick start
CODEBLOCK1
Operational guidance
- Keep input.contents as an array; malformed shapes are a common 400 cause. Pin the output dimension to match your index schema before writing vectors. Use the same model and dimension across one vector index to avoid mixed-vector incompatibility. For large image or video batches, stage files in object storage and reference stable URLs. Output location
- Default output: INLINECODE28 Override base dir with OUTPUT_DIR. References
技能名称: aliyun-qwen-multimodal-embedding
详细描述:
Model Studio 多模态嵌入
验证
bashcompile skills/ai/search/aliyun-qwen-multimodal-embedding/scripts/prepare multimodalembedding request.py && echo pycompile ok > output/aliyun-qwen-multimodal-embedding/validate.txt
通过标准: 命令退出码为0，且生成 output/aliyun-qwen-multimodal-embedding/validate.txt 文件。
输出与证据
- 将标准化后的请求负载、所选维度及样本输入参考保存至 output/aliyun-qwen-multimodal-embedding/ 目录下。 记录确切的模型、模态组合及输出向量维度，以确保可复现性。 当任务需要从 Model Studio 获取文本、图像或视频嵌入以用于检索或相似度工作流时，请使用此技能。
关键模型名称
根据需要，使用以下确切的模型字符串之一：
- qwen3-vl-embedding qwen2.5-vl-embedding tongyi-embedding-vision-plus-2026-03-06 选择指南：
- 对于最新的多模态嵌入路径，优先选择 qwen3-vl-embedding。 当需要与旧版部署管道兼容时，使用 qwen2.5-vl-embedding。 前提条件
- 在环境中设置 DASHSCOPEAPI KEY，或将 dashscopeapi key 添加到 ~/.alibabacloud/credentials 文件中。 在构建检索系统时，将此技能与向量存储（如 DashVector、OpenSearch 或 Milvus）配合使用。 标准化接口 (embedding.multimodal)
请求
- model (字符串，可选): 默认为 qwen3-vl-embedding texts (字符串数组，可选) images (字符串数组，可选): 公共URL或由客户端层上传的本地路径 videos (字符串数组，可选): 支持的公共URL dimension (整数，可选): 例如 qwen3-vl-embedding 支持 2560、2048、1536、1024、768、512、256 响应
- embeddings (对象数组) dimension (整数) usage (对象，可选) 快速开始
bashmultimodal embedding_request.py \
操作指南
- 保持 input.contents 为数组格式；格式错误是导致400错误的常见原因。 在写入向量之前，固定输出维度以匹配您的索引模式。 在同一个向量索引中使用相同的模型和维度，以避免混合向量不兼容问题。 对于大批量的图像或视频，将文件暂存于对象存储中，并引用稳定的URL。 输出位置
- 默认输出: output/aliyun-qwen-multimodal-embedding/request.json 可通过 OUTPUT_DIR 覆盖基础目录。 参考资料
