OpenAI: Text Embedding 3 Small (batch)

OpenAI: Text Embedding 3 Small (batch) is an embedding model from OpenAI, offered through api14. It belongs to the embeddings category, and its model id on api14 is openai/text-embedding-3-small:batch. OpenAI describes text-embedding-3-small as an improved, more performant version of the ada embedding model. Embeddings are a numerical representation of text that can be used to measure the relatedness between two pieces of text.

The model takes text as its only input and returns embeddings as its only output. It has a context length of 8192 tokens, and the maximum output is 7372. There are no supported parameters listed for this model, so you cannot tune sampling or generation settings through the request. The batch suffix in the model id indicates the batch variant of this embedding model on api14.

Because it produces embeddings rather than generated text, the practical uses follow from measuring similarity between texts. Typical work includes semantic search, where you embed a query and a set of documents and rank documents by closeness; retrieval for a retrieval augmented generation pipeline, where embeddings select relevant passages to pass to a chat model; clustering or topic grouping of large text collections; deduplication, where near identical items are found by comparing vectors; and classification or routing, where a lightweight model consumes the embeddings. The 8192 token context means individual inputs can be fairly long, so whole pages or long passages can be embedded without splitting them first.

To call it through api14, use the OpenAI-compatible API with the model id openai/text-embedding-3-small:batch. Point your existing OpenAI client at the api14 base URL and send your text to the embeddings endpoint, or use the equivalent chat style route if your client expects one. Because the model lists no supported parameters, keep requests simple: send the model id and the input text, and read the embedding vector from the response. There is no need to set temperature, top_p, or similar options, since this model does not accept them.

One thing to keep in mind is that this model only handles text input and only returns embeddings. It does not generate chat completions or any other text output, so if your task is conversation, summarization, or code generation, a chat or completion model is the better fit. If you need to embed very large volumes of text, the batch variant is the relevant choice here, while other embedding models may suit different accuracy or language needs.