text-embedding-3-large is the high-precision variant among OpenAI's third-generation text embedding models. It converts natural language and code into numerical vectors for semantic search, knowledge base retrieval, and text clustering. It provides full representations of up to 3072 dimensions and also supports reduced dimensions, making it suitable for targeted trade-offs between retrieval quality and index size, rather than directly generating chat answers.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
Input parameters and result formats vary by service. Use the public API for this model and follow its guide for generation, task retrieval and editing operations.
Clarify capacity, inputs and outputs, and invocation methods before selecting a model.
Vector dimensions are native model specifications; input organization, batch item count, and return encoding fall within the scope of this platform's embedding interface.
Learn what text-embedding-3-large can bring to your work.
The model converts text content into computable conceptual representations, so retrieval does not need to rely solely on whether keywords match exactly. Questions, passages, and code descriptions can each be converted into vectors, then compared for similarity by a retrieval system. It provides the representations, while document ranking, filtering, and final answers are handled by the application workflow.
In multilingual retrieval and English-task evaluations at release, large achieved higher average scores than the same-generation small and the previous-generation ada-002. For multilingual knowledge bases, queries with diverse phrasing, or content that requires distinguishing between closely related topics, it can be considered when quality is the priority, with results then validated using real queries.
Full vectors provide representations of up to 3072 dimensions, and output can also be shortened through dimensions. Shorter vectors can reduce the burden of vector storage and similarity computation, but may sacrifice some accuracy. When an existing vector database has dimensional constraints, test a shortened configuration first instead of immediately replacing the entire embedding model.
Start with specific tasks to find where the model can make an impact.
Split product manuals, operating procedures, and frequently asked questions into text passages, generate vectors in batches, and save their corresponding document locations. When users ask questions, generate vectors for the questions using the same model and dimensions, retrieve relevant passages, and pass them to an answer model. The deliverable is a searchable content index, not answers written directly by the embedding API.
For help centers or research repositories containing materials in different languages, convert titles, summaries, and body excerpts into vectors to establish a foundation for semantic retrieval. Before launch, prepare representative queries for each language and check whether relevant content can enter the candidate results; then combine filters for language, product, and time to create a usable search list.
Input ticket descriptions, user feedback, or code explanations to obtain a corresponding vector for each item, for subsequent clustering and similarity analysis. This can help identify feedback with similar topics, potential duplicate content, and related technical explanations. Final category labels and duplicate determinations should be confirmed by business rules or manual spot checks, rather than by directly reading vector values.
Choose based on task complexity, input materials, and expected results.
text-embedding-3-small is more geared toward efficient applications, while large is better suited for tasks that prioritize retrieval quality. Do not replace all indexes simply because the model is larger: first fix the corpus, queries, and retrieval workflow, compare relevant-document recall, then evaluate vector size and computational burden. If small already meets requirements, keep it; if difficult cases remain common, then test large.
Compared with text-embedding-ada-002, large offers stronger published benchmark performance and native dimension-shortening capability. The official documentation also provides an example where large's 256-dimensional representation exceeds full ada-002 on MTEB. When migrating, regenerate document and query vectors, and standardize the model and dimensions; do not merely change the query-side model while continuing to use the old index.
Prepare the inputs first, then connect them to the corresponding application workflow.
Select Chinese and English document excerpts and real queries, retaining a document ID, title, and access permissions for each excerpt.
Use /v1/embeddings to select text-embedding-3-large, and manage the text and its document ID separately. Write vectors to an index with a fixed configuration, and use the same encoding for queries; first confirm that candidate source text is traceable, then use retrieval results in the application.
Design the task directly from the inputs and acceptance priorities below.
Use the same model and dimensions to generate vectors for documents and queries, then perform similarity retrieval; compare full vectors with shortened-dimension configurations.
Calculate recall for manually labeled relevant passages, and check cross-lingual and synonymous expressions; retain original sources in the results, while a downstream model is responsible for generating answers.
Before formal use, understand the output quality and capability scope.
Answers to common questions about using text-embedding-3-large.
It cannot directly generate answers. It converts questions and document chunks into vectors, helping the retrieval system find relevant content. A complete question-answering workflow typically retrieves relevant passages first, then uses a generation model to compose the answer; the embedding returned by the embeddings API is not itself an answer and does not include an explanation.
When quality is the priority, you can start with the full dimensionality; if your vector database has dimension limits or the index is large, you can test a reduced-dimension configuration. Official examples provide 1024- and 256-dimension options, but the specific trade-off should be determined using your own query set after comparing retrieval performance and storage overhead.
At release, large performed better on average in multilingual retrieval and English task evaluations, making it suitable for applications with higher retrieval quality requirements. small is oriented toward efficiency. It is recommended to compare them using the same corpus and real queries, focusing on difficult cases rather than deciding based solely on the model name.
You can place multiple non-empty texts in the input array to generate multiple vectors in a single request. In the response data, each item includes an index and embedding, which can be used to associate it with the original input. When saving vectors, you should also save the document identifier, chunk position, and dimensions used to facilitate later retrieval.
float returns an array of numbers by default, suitable for direct use with programs or vector databases that require numeric vectors. base64 returns an encoded string, suitable for systems that already have a corresponding decoding workflow. They change the return representation, not the embedding task, and do not replace dimensionality control through dimensions.
Choose a development approach or application suited to this model and follow its setup guide.
↗Python SDK Integration Guide
SDK installation, authentication and examples↗TypeScript SDK Integration Guide
SDK installation, authentication and examples
↗Input
$0.0173per 1M tokens
Official $0.13
You pay$13,584for 142,800 credits · $0.0951 / credits46% OFF
You're billed in credits. Pick a top-up tier to see its per-credit rate — the larger the top-up, the cheaper each credit.
This is only a basic call example. See the full docs for more parameters and advanced usage.
View Docs