BAAI: bge-base-en-v1.5

BAAI: bge-base-en-v1.5 is an embedding model from BAAI, the Beijing Academy of Artificial Intelligence. It belongs to the embeddings category, which means it is not built to generate chat replies or complete text. Instead, it turns text into numerical vectors. It is an English-language model, and the v1.5 in its name marks this as the 1.5 version of the bge-base-en family. If your application needs a general chat assistant, this is not that kind of model. If your application needs to compare, search or match text by meaning, this is the kind of model to evaluate.

The model accepts text as input and produces embeddings as output. Specifically, it converts English sentences and paragraphs into 768-dimensional dense vectors, which are intended to carry semantic meaning. The context length is 512 tokens, and the maximum output is 460. In practice, 512 tokens means you should send sentences or single paragraphs rather than long documents, since longer material will need to be split before embedding. The provider describes the model as delivering efficient, high-quality semantic embeddings optimized for retrieval, semantic search, and document-matching workflows. The vectors are dense, so they are compact enough to store and compare at scale.

Those characteristics point to a fairly specific set of use cases. Semantic search is the clearest one: embed a query and your documents, then rank documents by vector similarity instead of keyword overlap. Retrieval pipelines can use the same behaviour to fetch passages that are relevant to a question. Document matching fits as well, for example finding near-duplicate records, grouping similar support tickets, or linking related items across two collections. Because the model is English-only and limited to 512 tokens, it suits sentence and paragraph level tasks. Deduplication, clustering and similarity scoring on short English text are all natural fits, provided you chunk longer inputs first.

You call the model through api14's OpenAI-compatible API using the model id "baai/bge-base-en-v1.5". Because the API follows the OpenAI format, you can point an existing OpenAI client at api14, send your text to the embeddings endpoint, and receive the vectors in the response. The listed supported parameters for this model are empty, so it takes no extra tuning options beyond what your request already includes; there is nothing like temperature or top_p to set, since those belong to text generation rather than embedding. Check the live price table on this page for current rates.

The main constraints to keep in mind are the 512 token context length, the 460 maximum output, English-only input, and the fact that the output is vectors rather than prose. If you need to embed long documents in one pass, handle multiple languages, or generate text, another model on api14 will fit better. Within its intended scope, retrieval, semantic search and document matching on English sentences and paragraphs, this model is a direct option.