Perform inference on the service | Elasticsearch API documentation (v8)

Perform inference on the service Generally available; Added in 8.11.0

POST /_inference/{inference_id}

This API enables you to use machine learning models to perform specific tasks on data that you provide as an input. It returns a response with the results of the tasks. The inference endpoint you use can perform one specific task that has been defined when the endpoint was created with the create inference API.

For details about using this API with a service, such as Amazon Bedrock, Anthropic, or HuggingFace, refer to the service-specific documentation.

The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Azure, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face. For built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.

Required authorization

Cluster privileges: monitor_inference

Path parameters

inference_id string Required

The unique identifier for the inference endpoint.

Query parameters

timeout string

The amount of time to wait for the inference request to complete.

Values are -1 or 0.

application/json

Body

query string

The query input, which is required only for the rerank task. It is not required for other tasks.
input string | array[string] Required

The text on which you want to perform the inference task. It can be a single string or an array.

Inference endpoints for the completion task type currently only support a single string as input.

One of:
string-1 string array-2 array[string]
task_settings object

Responses

200 application/json
Hide response attributes Show response attributes object
- text_embedding_bytes array[object]
  
  The text embedding result object for byte representation
  
  Hide text_embedding_bytes attribute Show text_embedding_bytes attribute object
  
  embedding array[number] Required
  
  Text Embedding results containing bytes are represented as Dense Vectors of bytes.
- text_embedding_bits array[object]
  
  The text embedding result object for byte representation
  
  Hide text_embedding_bits attribute Show text_embedding_bits attribute object
  
  embedding array[number] Required
  
  Text Embedding results containing bytes are represented as Dense Vectors of bytes.
- text_embedding array[object]
  
  The text embedding result object
  
  Hide text_embedding attribute Show text_embedding attribute object
  
  embedding array[number] Required
  
  Text Embedding results are represented as Dense Vectors of floats.
- sparse_embedding array[object]
  
  Hide sparse_embedding attribute Show sparse_embedding attribute object
  
  embedding object Required
  
  Sparse Embedding tokens are represented as a dictionary of string to double.
  
  Hide embedding attribute Show embedding attribute object
  
  * number Additional properties
- completion array[object]
  
  The completion result object
  
  Hide completion attribute Show completion attribute object
  
  result string Required
- rerank array[object]
  
  The rerank result object representing a single ranked document id: the original index of the document in the request relevance_score: the relevance_score of the document relative to the query text: Optional, the text of the document, if requested
  
  Hide rerank attributes Show rerank attributes object
  
  index number Required
  
  relevance_score number Required
  
  text string