Create Pipeline

Create a new named pipeline with custom chunking and refinement steps

Create Pipeline

cURL

curl --request POST \
  --url https://api.example.com/v1/pipeline

200

{
  "id": "<string>",
  "slug": "<string>",
  "description": "<string>",
  "organization_slug": "<string>",
  "steps": [
    {}
  ],
  "created_at": "<string>",
  "updated_at": "<string>"
}

Create a reusable pipeline configuration that can be executed multiple times with different inputs.

Examples

Python

from chonkie.cloud import Pipeline

pipeline = Pipeline(
    slug="my-rag-pipeline",
    description="Pipeline for RAG processing"
).chunk_with("recursive", chunk_size=512).refine_with("overlap", context_size=64)

# Pipeline is auto-saved on first run
chunks = pipeline.run(text="Your document text here")

Request

slug · string · required

Unique identifier for the pipeline. Must contain only lowercase letters, numbers, dashes, and underscores.

description · string

Optional description of the pipeline’s purpose.

steps · array · required

Array of pipeline step configurations. Must include at least one chunk step.

Each step object contains:

• type: Step type (process, chunk, or refine)
• component: Component name (e.g., recursive, overlap, embeddings)
• Additional parameters specific to the component

Response

id · string

Unique identifier for the pipeline.

slug · string

The pipeline slug.

description · string

Pipeline description.

organization_slug · string

Organization that owns the pipeline.

steps · array

Array of pipeline step configurations.

created_at · string

ISO 8601 timestamp of creation.

updated_at · string

ISO 8601 timestamp of last update.

Errors

Status	Description
409	Pipeline with this slug already exists
422	Invalid slug format or missing required fields