Text Summarization — batch summaries for up to 10 texts

Batch Text Summarization API

What it does

Text Summarization batches up to 10 input texts and returns a summary for each one in a single GPU pass. Send plain text strings in texts, choose model as bart or pegasus, and optionally set max_length and min_length to control summary size.

The response includes the total count, the model used, and a summaries array. Each item contains the generated summary plus input_length and summary_length, so you can compare source length with the output you received.

Use Text Summarization when you need to condense articles, support tickets, research notes, or user feedback into shorter versions for review, search, or downstream AI workflows. Because it accepts multiple texts at once, it fits batch processing jobs where latency and throughput both matter.

If you are building a content pipeline, you can summarize many records in one request and keep the output structured enough to store, rank, or pass into another model.

Quickstart

Summarize one or more texts with the default model.

curl -X POST "http://localhost:8080/dosvak/batch-summarize-text" \
  -H "apy-token: $APY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "texts": [
      "ApyHub provides simple APIs for common developer tasks. This endpoint turns longer text into shorter summaries."
    ]
  }'

What you'll get back

Returns a JSON object with count integer, model string, and summaries array. Each item in summaries is an object with summary string, input_length integer, and summary_length integer.

{
  "count": 1,
  "model": "bart",
  "summaries": [
    {
      "summary": "ApyHub provides simple APIs for common developer tasks.",
      "input_length": 110,
      "summary_length": 55
    }
  ]
}

About this endpoint

What it does

Summarizes up to 10 input texts in a single request and returns one summary per text. You can choose the summarization model and control the generated summary lengths.

Request Body

Parameter	Type	Mandatory	Description
texts	String Array	Yes	The texts to summarize. Minimum 1 item, maximum 10 items.
model	ENUM	No	Summarization model to use. Allowed values: `bart`, `pegasus`. Default: `bart`.
max_length	Integer	No	Maximum length of each generated summary. Default: 200. Minimum: 30, maximum: 512.
min_length	Integer	No	Minimum length of each generated summary. Default: 40. Minimum: 10, maximum: 200.

Response

Returns a JSON object with count as an integer, model as a string, and summaries as an array of objects. Each object in summaries contains a summary string, input_length integer, and summary_length integer.

Parameter	Type	Mandatory	Description
count	Integer	No	The number of summaries returned.
model	String	No	The model used to generate the summaries.
summaries	Object Array	No	An array of summary result objects.
summaries[].summary	String	No	The generated summary text.
summaries[].input_length	Integer	No	The length of the corresponding input text.
summaries[].summary_length	Integer	No	The length of the generated summary.

Body

Name

Type

Description

bodyREQUIRED

object

—

Quickstart

What you'll get back

About this endpoint

What it does

Request Body

Response

Body

Errors any endpoint can return