Skip to main content
Late Chunker
curl --request POST \
  --url https://api.chonkie.ai/v1/chunk/late
{
  "text": "<string>",
  "start_index": 123,
  "end_index": 123,
  "token_count": 123
}
The Late Chunker generates contextual embeddings for each token before chunking, enabling more semantically aware chunk boundaries.

​
Examples

​
Text Input

from chonkie.cloud import LateChunker

chunker = LateChunker(
    embedding_model="minishlab/potion-base-8M",
    chunk_size=512
)

text = "Your text here..."
chunks = chunker.chunk(text)

​
File Input

from chonkie.cloud import LateChunker

chunker = LateChunker(
    embedding_model="minishlab/potion-base-8M",
    chunk_size=512
)

# Chunk from file
with open("document.txt", "rb") as f:
    chunks = chunker.chunk(file=f)

​
Request

​
Parameters

​
text
string | string[]
The text to chunk. Can be a single string or an array of strings for batch processing. Either text or file is required.
​
file
file
File to chunk. Use multipart/form-data encoding. Either text or file is required.
​
embedding_model
string
default:"minishlab/potion-base-8M"
The embedding model to use.
Chosen embedding model must return token level embeddings. Request will fail otherwise
​
chunk_size
integer
default:"512"
Maximum number of tokens per chunk.
​
recipe
string
default:"default"
Recursive rules recipe to follow. See all recipes on our Hugging Face repo
​
lang
string
default:"en"
Language of the document to chunk

​
Response

​
Returns

Array of Chunk objects, each containing:
​
text
string
The chunk text content.
​
start_index
integer
Starting character position in the original text.
​
end_index
integer
Ending character position in the original text.
​
token_count
integer
Number of tokens in the chunk.
⌘I