GTE
An open-source series of text embedding models released by Alibaba. GTE models support text vectorization for use cases such as retrieval-augmented generation (RAG) and are available as free open-source releases.
Timeline (last 90 days)
Embedding models are neural network models that map discrete inputs into dense vectors, with representative examples including OpenAI's text-embedding-ada-002, Voyage AI's voyage series, and open-source BGE, E5, GTE and others.
GTE comes from the paper "mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval" and is a BERT-style bidirectional encoder that replaces absolute position embeddings with RoPE, uses gated MLP, and applies layer normalization after each residual connection.
The GTE architecture underpins Alibaba's gte-*-v1.5 and gte-multilingual-* series as well as Snowflake's snowflake-arctic-embed-m-v2.0 checkpoints.
OpenAI's embeddings are paid, DeepSeek's embeddings are not open-source and can only be accessed by purchasing tokens, and Alibaba has open-sourced its GTE series of embedding models.
All Facts (4)
Embedding models are neural network models that map discrete inputs into dense vectors, with representative examples including OpenAI's text-embedding-ada-002, Voyage AI's voyage series, and open-source BGE, E5, GTE and others.
50%UnverifiedGTE comes from the paper "mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval" and is a BERT-style bidirectional encoder that replaces absolute position embeddings with RoPE, uses gated MLP, and applies layer normalization after each residual connection.
50%UnverifiedThe GTE architecture underpins Alibaba's gte-*-v1.5 and gte-multilingual-* series as well as Snowflake's snowflake-arctic-embed-m-v2.0 checkpoints.
50%UnverifiedOpenAI's embeddings are paid, DeepSeek's embeddings are not open-source and can only be accessed by purchasing tokens, and Alibaba has open-sourced its GTE series of embedding models.
50%