Gemini Embedding Text Embedding Model
Gemini Embedding is an advanced text embedding model that provides powerful language understanding capabilities through the Gemini API.
0
- Provide high-precision text embedding to capture semantics and context
- Supports multilingual text processing for over 100 languages
- 8K input marker length, capable of processing long text and code
- 3K output dimension, providing high-precision semantic representation
- Nested Representation Learning (MRL), flexibly adjusting dimensions to optimize storage and performance
Product Details
Gemini Embedding is an experimental text embedding model launched by Google, providing services through the Gemini API. This model performs excellently in the Multi Language Text Embedding Benchmark (MTEB), surpassing previous state-of-the-art models. It can convert text into high-dimensional numerical vectors, capture semantic and contextual information, and is widely used in scenarios such as retrieval, classification, and similarity detection. Gemini Embedding supports over 100 languages, with 8K input tag lengths and 3K output dimensions, and introduces Nested Representation Learning (MRL) technology to flexibly adjust dimensions to meet storage requirements. The model is currently in the experimental stage and a stable version will be released in the future.
