Query Completion

Generate completion using relevant chunks as context.

POST /query

JavaScript

JavaScript

import Morphik from 'morphik';

const client = new Morphik({
  apiKey: process.env['MORPHIK_API_KEY'], // This is the default and can be omitted
});

const response = await client.query.generateCompletion();

console.log(response.completion);

200

{
  "completion": "<string>",
  "usage": {},
  "finish_reason": "<string>",
  "sources": [],
  "metadata": {}
}

Headers

authorization · string | null

Body

application/json

Request model for completion generation

query · string | null

Natural-language query used to retrieve relevant chunks or documents.

Minimum string length: 1

query_image · string | null

Base64-encoded image to use as query for Morphik multimodal retrieval. Requires use_colpali=True. Mutually exclusive with 'query'.

filters · Filters · object

Metadata filters supporting logical operators ($and/$or/$not/$nor) and field predicates ($eq/$ne/$gt/$gte/$lt/$lte/$in/$nin/$exists/$type/$regex/$contains).

k · integer · default: 4

Maximum number of chunks or documents to return.

min_score · number · default: 0

Minimum similarity score a result must meet before it is returned.

use_reranking · boolean | null

When provided, overrides the workspace reranking configuration for this request.

use_colpali · boolean | null

When provided, uses Morphik's finetuned ColPali style embeddings (recommended to be True for high quality retrieval).

output_format · enum<string> | null · default: base64

How to return image chunks: base64 (default), url, or text (markdown format)

Available options: base64, url, text

padding · integer · default: 0

Number of additional chunks/pages to retrieve before and after matched chunks (ColPali only)

Required range: x >= 0

folder_name · string string[]

Optional folder scope. Accepts a folder PATH (e.g., '/Company/Reports') or list of paths.

folder_depth · integer | null

Folder scope depth. 0 or None = exact folder only, -1 = include all descendants, n > 0 = include descendants up to n levels deeper.

end_user_id · string | null

Optional end-user scope for the operation

max_tokens · integer | null

Maximum number of tokens allowed in the generated completion.

temperature · number | null

Sampling temperature passed to the completion model (None uses provider default).

prompt_overrides · QueryPromptOverrides · object

Optional customizations for entity extraction, resolution, and query prompts

Show child attributes

schema · any Schema · object

Schema for structured output, can be a Pydantic model or JSON schema dict

chat_id · string | null

Optional chat session ID for persisting conversation history

stream_response · boolean | null · default: false

Whether to stream the response back in chunks

llm_config · Llm Config · object

LiteLLM-compatible model configuration (e.g., model name, API key, base URL)

inline_citations · boolean | null · default: false

Whether to include inline citations with filename and page number in the response

Response

200

application/json

Successful Response

Response from completion generation

completion · string StructuredCompletion · object · required

usage · Usage · object · required

Show child attributes

finish_reason · string | null

sources · ChunkSource · object[]

Show child attributes

metadata · Metadata · object