Alibaba Cloud - API - verified
Qwen3-VL Embedding embeddings
Qwen3-VL Embedding is a current embedding model from Alibaba Cloud, called by API as qwen3-vl-embedding. It returns 2,560-dimension vectors by default and accepts up to 32,000 input tokens, for text, image and video input. The first-party list price is $0.10 per 1M input tokens.
CurrentModel id: qwen3-vl-embedding
- Default dims
- 2,560
- Max input
- 32,000 tok
- Price / 1M
- $0.10
- Inputs
- Text, Image, Video
Dimensions
Flexible output size- 2,560default
- 2,048
- 1,536
- 1,024
- 768
- 512
- 256
Default output: 2,560 dimensions.
Listed sizes: 2,560, 2,048, 1,536, 1,024, 768, 512 and 256.
Alibaba Cloud documents shorter output vectors for this model (Matryoshka-style), so you can store smaller embeddings by asking for fewer dimensions instead of re-embedding with another model.
Pricing
- Per 1M input tokens
- $0.10
- Batch, per 1M
- Not stated
China (Beijing) text price; image and video input is $0.258 per 1M tokens, with 1 million free tokens for 90 days.
Usage
Copied verbatim from the page linked below. Placeholders such as project ids and API keys are the source's own.
# Use qwen3-vl-embedding to generate a fused vector.
resp = dashscope.MultiModalEmbedding.call(
# If you have not set an environment variable, replace the following line with your Model Studio API key: api_key="sk-xxx",
api_key=os.getenv("DASHSCOPE_API_KEY"),
model="qwen3-vl-embedding",
input=input_data,
enable_fusion=True,
# Optional parameter: Specifies the vector dimension. Supported values: 2560, 2048, 1536, 1024, 768, 512, and 256. Default: 2560.
# dimension = 1024
)
print(json.dumps(resp.output, indent=4))Worth knowing
Generates fused or independent vectors from text, one image and video for cross-modal retrieval and image search, supporting 33 languages and up to 20 content elements per request.
Release date not stated in the sources. Multilingual: yes.
FAQ
How many dimensions does Qwen3-VL Embedding output?
Default output: 2,560 dimensions. Listed sizes: 2,560, 2,048, 1,536, 1,024, 768, 512 and 256. Alibaba Cloud documents shorter output vectors for this model (Matryoshka-style), so you can store smaller embeddings by asking for fewer dimensions instead of re-embedding with another model.
What is the max input length of Qwen3-VL Embedding?
32,000 tokens per input, per Alibaba Cloud's documentation. Longer text has to be chunked before you embed it.
How much does Qwen3-VL Embedding cost?
The first-party list price is $0.10 per 1M input tokens. China (Beijing) text price; image and video input is $0.258 per 1M tokens, with 1 million free tokens for 90 days. Source: https://www.alibabacloud.com/help/en/model-studio/model-pricing.
What can Qwen3-VL Embedding embed?
Text, image and video input, per the provider. It is documented as multilingual.
Sources
Last verified from the pages below.
- Model Studio embedding docs - alibabacloud.com
- Model Studio pricing - alibabacloud.com
More Alibaba Cloud embedding models
Default dimensions, max input tokens and list price per 1M tokens.
- Qwen3.7 Text Embedding1,024 dims - 128,000 tokens - $0.07
- Qwen3.7 Text Embedding Flashdims not stated - 128,000 tokens - Not stated
- Text Embedding V41,024 dims - 8,192 tokens - $0.07
- Text Embedding V31,024 dims - 8,192 tokens - $0.07
- Tongyi Embedding Vision Plus1,152 dims - 1,024 tokens - $0.09
- Tongyi Embedding Vision Flash768 dims - 1,024 tokens - $0.09
- Multimodal Embedding V11,024 dims - 512 tokens - Not stated
Newsletter
New models, every week
One email per week. New model releases, embedding and retrieval updates and what they mean for what you build.
Keep exploring
Model API pricing and specs
First-party prices per 1M tokens for chat and reasoning models.
Local models
Open-weight LLMs you can download and run, with sizes and licenses.
What is RAG?
Where embeddings fit in a retrieval pipeline, end to end.
Vector databases compared
Where the vectors go once you have picked a model.