Multilingual E5 Small

Multilingual E5 Small is part of the E5 family of text embedding models. This small variant contains 12 layers. This model is well-suited for tasks such as:

Semantic Search: Finding documents or text passages that are semantically relevant to a query.
Text Clustering: Grouping similar pieces of text based on their semantic meaning.
Informative Response Generation: Following instructions to provide summaries, explanations, or answers to factual queries.

View in Model Garden View pricing

Model ID	`multilingual-e5-small-maas`
Modalities	Text Input only Image Not supported Audio Not supported Video Not supported Embeddings Output only
Output dimensions	Up to 384
Number of layers	12
Maximum sequence length	512 tokens
Consumption options	Provisioned Throughput Not supported Batch inference Not supported Pay-as-you-go Standard PayGo Supported Fixed quota Supported
Supported regions	Model availability	United States: `us-central1` Europe: `europe-west4`
Supported regions	ML processing	Multi-region: `us`, `eu`
Quotas	`us-central1`: 25,200 tokens per minute `europe-west4`: 25,200 tokens per minute
Versions	`multilingual-e5-small` Launch stage: GA Release date: September 19, 2025
Supported languages	See Supported languages.

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2026-07-23 UTC.