Перейти к основному содержимому

REST API

API docs by Redocly

Compressa API (1.0.0)

Download OpenAPI specification:Download

Chat API

Compressa Chat API is OpenAI compatible API that allows you to create chat completions.

Create Chat Completion

Request Body schema: application/json
required
required
Array of ChatCompletionSystemMessageParam (object) or ChatCompletionUserMessageParam (object) or ChatCompletionAssistantMessageParam (object) or ChatCompletionToolMessageParam (object) or ChatCompletionFunctionMessageParam (object) or CustomChatCompletionMessageParam (object) (Messages)
model
required
string (Model)
Frequency Penalty (number) or Frequency Penalty (null) (Frequency Penalty)
Default: 0
Logit Bias (object) or Logit Bias (null) (Logit Bias)
Logprobs (boolean) or Logprobs (null) (Logprobs)
Default: false
Top Logprobs (integer) or Top Logprobs (null) (Top Logprobs)
Default: 0
Max Tokens (integer) or Max Tokens (null) (Max Tokens)
N (integer) or N (null) (N)
Default: 1
Presence Penalty (number) or Presence Penalty (null) (Presence Penalty)
Default: 0
ResponseFormat (object) or null
Seed (integer) or Seed (null) (Seed)
Stop (string) or Array of Stop (strings) or Stop (null) (Stop)
Stream (boolean) or Stream (null) (Stream)
Default: false
StreamOptions (object) or null
Temperature (number) or Temperature (null) (Temperature)
Default: 0.7
Top P (number) or Top P (null) (Top P)
Default: 1
Array of Tools (objects) or Tools (null) (Tools)
"none" (string) or ChatCompletionNamedToolChoiceParam (object) or Tool Choice (null) (Tool Choice)
Default: "none"
User (string) or User (null) (User)
Best Of (integer) or Best Of (null) (Best Of)
Use Beam Search (boolean) or Use Beam Search (null) (Use Beam Search)
Default: false
Top K (integer) or Top K (null) (Top K)
Default: -1
Min P (number) or Min P (null) (Min P)
Default: 0
Repetition Penalty (number) or Repetition Penalty (null) (Repetition Penalty)
Default: 1
Length Penalty (number) or Length Penalty (null) (Length Penalty)
Default: 1
Early Stopping (boolean) or Early Stopping (null) (Early Stopping)
Default: false
Ignore Eos (boolean) or Ignore Eos (null) (Ignore Eos)
Default: false
Min Tokens (integer) or Min Tokens (null) (Min Tokens)
Default: 0
Array of Stop Token Ids (integers) or Stop Token Ids (null) (Stop Token Ids)
Skip Special Tokens (boolean) or Skip Special Tokens (null) (Skip Special Tokens)
Default: true
Spaces Between Special Tokens (boolean) or Spaces Between Special Tokens (null) (Spaces Between Special Tokens)
Default: true
Echo (boolean) or Echo (null) (Echo)
Default: false

If true, the new message will be prepended with the last message if they belong to the same role.

Add Generation Prompt (boolean) or Add Generation Prompt (null) (Add Generation Prompt)
Default: true

If true, the generation prompt will be added to the chat template. This is a parameter used by chat template in tokenizer config of the model.

Add Special Tokens (boolean) or Add Special Tokens (null) (Add Special Tokens)
Default: false

If true, special tokens (e.g. BOS) will be added to the prompt on top of what is added by the chat template. For most models, the chat template takes care of adding the special tokens so this should be set to False (as is the default).

Include Stop Str In Output (boolean) or Include Stop Str In Output (null) (Include Stop Str In Output)
Default: false

Whether to include the stop string in the output. This is only applied when the stop or stop_token_ids is set.

Guided Json (string) or Guided Json (object) or BaseModel (object) or Guided Json (null) (Guided Json)

If specified, the output will follow the JSON schema.

Guided Regex (string) or Guided Regex (null) (Guided Regex)

If specified, the output will follow the regex pattern.

Array of Guided Choice (strings) or Guided Choice (null) (Guided Choice)

If specified, the output will be exactly one of the choices.

Guided Grammar (string) or Guided Grammar (null) (Guided Grammar)

If specified, the output will follow the context free grammar.

Guided Decoding Backend (string) or Guided Decoding Backend (null) (Guided Decoding Backend)

If specified, will override the default guided decoding backend of the server for this specific request. If set, must be either 'outlines' / 'lm-format-enforcer'

Guided Whitespace Pattern (string) or Guided Whitespace Pattern (null) (Guided Whitespace Pattern)

If specified, will override the default whitespace pattern for guided json decoding.

Enforced Str (string) or Enforced Str (null) (Enforced Str)

Responses

Request samples

Content type
application/json
{
  • "messages": [
    ],
  • "model": "string",
  • "frequency_penalty": 0,
  • "logit_bias": {
    },
  • "logprobs": false,
  • "top_logprobs": 0,
  • "max_tokens": 0,
  • "n": 1,
  • "presence_penalty": 0,
  • "response_format": {
    },
  • "seed": -9223372036854776000,
  • "stop": "string",
  • "stream": false,
  • "stream_options": {
    },
  • "temperature": 0.7,
  • "top_p": 1,
  • "tools": [
    ],
  • "tool_choice": "none",
  • "user": "string",
  • "best_of": 0,
  • "use_beam_search": false,
  • "top_k": -1,
  • "min_p": 0,
  • "repetition_penalty": 1,
  • "length_penalty": 1,
  • "early_stopping": false,
  • "ignore_eos": false,
  • "min_tokens": 0,
  • "stop_token_ids": [
    ],
  • "skip_special_tokens": true,
  • "spaces_between_special_tokens": true,
  • "echo": false,
  • "add_generation_prompt": true,
  • "add_special_tokens": false,
  • "include_stop_str_in_output": false,
  • "guided_json": "string",
  • "guided_regex": "string",
  • "guided_choice": [
    ],
  • "guided_grammar": "string",
  • "guided_decoding_backend": "string",
  • "guided_whitespace_pattern": "string",
  • "enforced_str": "string"
}

Response samples

Content type
application/json
null

Create Completion

Request Body schema: application/json
required
model
required
string (Model)
required
Array of Prompt (integers) or Array of Prompt (integers) or Prompt (string) or Array of Prompt (strings) (Prompt)
Best Of (integer) or Best Of (null) (Best Of)
Echo (boolean) or Echo (null) (Echo)
Default: false
Frequency Penalty (number) or Frequency Penalty (null) (Frequency Penalty)
Default: 0
Logit Bias (object) or Logit Bias (null) (Logit Bias)
Logprobs (integer) or Logprobs (null) (Logprobs)
Max Tokens (integer) or Max Tokens (null) (Max Tokens)
Default: 16
n
integer (N)
Default: 1
Presence Penalty (number) or Presence Penalty (null) (Presence Penalty)
Default: 0
Seed (integer) or Seed (null) (Seed)
Stop (string) or Array of Stop (strings) or Stop (null) (Stop)
Stream (boolean) or Stream (null) (Stream)
Default: false
StreamOptions (object) or null
Suffix (string) or Suffix (null) (Suffix)
Temperature (number) or Temperature (null) (Temperature)
Default: 1
Top P (number) or Top P (null) (Top P)
Default: 1
User (string) or User (null) (User)
Use Beam Search (boolean) or Use Beam Search (null) (Use Beam Search)
Default: false
Top K (integer) or Top K (null) (Top K)
Default: -1
Min P (number) or Min P (null) (Min P)
Default: 0
Repetition Penalty (number) or Repetition Penalty (null) (Repetition Penalty)
Default: 1
Length Penalty (number) or Length Penalty (null) (Length Penalty)
Default: 1
Early Stopping (boolean) or Early Stopping (null) (Early Stopping)
Default: false
Array of Stop Token Ids (integers) or Stop Token Ids (null) (Stop Token Ids)
Ignore Eos (boolean) or Ignore Eos (null) (Ignore Eos)
Default: false
Min Tokens (integer) or Min Tokens (null) (Min Tokens)
Default: 0
Skip Special Tokens (boolean) or Skip Special Tokens (null) (Skip Special Tokens)
Default: true
Spaces Between Special Tokens (boolean) or Spaces Between Special Tokens (null) (Spaces Between Special Tokens)
Default: true
Truncate Prompt Tokens (integer) or Truncate Prompt Tokens (null) (Truncate Prompt Tokens)
Include Stop Str In Output (boolean) or Include Stop Str In Output (null) (Include Stop Str In Output)
Default: false

Whether to include the stop string in the output. This is only applied when the stop or stop_token_ids is set.

ResponseFormat (object) or null

Similar to chat completion, this parameter specifies the format of output. Only {'type': 'json_object'} or {'type': 'text' } is supported.

Guided Json (string) or Guided Json (object) or BaseModel (object) or Guided Json (null) (Guided Json)

If specified, the output will follow the JSON schema.

Guided Regex (string) or Guided Regex (null) (Guided Regex)

If specified, the output will follow the regex pattern.

Array of Guided Choice (strings) or Guided Choice (null) (Guided Choice)

If specified, the output will be exactly one of the choices.

Guided Grammar (string) or Guided Grammar (null) (Guided Grammar)

If specified, the output will follow the context free grammar.

Guided Decoding Backend (string) or Guided Decoding Backend (null) (Guided Decoding Backend)

If specified, will override the default guided decoding backend of the server for this specific request. If set, must be one of 'outlines' / 'lm-format-enforcer'

Guided Whitespace Pattern (string) or Guided Whitespace Pattern (null) (Guided Whitespace Pattern)

If specified, will override the default whitespace pattern for guided json decoding.

Responses

Request samples

Content type
application/json
{
  • "model": "string",
  • "prompt": [
    ],
  • "best_of": 0,
  • "echo": false,
  • "frequency_penalty": 0,
  • "logit_bias": {
    },
  • "logprobs": 0,
  • "max_tokens": 16,
  • "n": 1,
  • "presence_penalty": 0,
  • "seed": -9223372036854776000,
  • "stop": "string",
  • "stream": false,
  • "stream_options": {
    },
  • "suffix": "string",
  • "temperature": 1,
  • "top_p": 1,
  • "user": "string",
  • "use_beam_search": false,
  • "top_k": -1,
  • "min_p": 0,
  • "repetition_penalty": 1,
  • "length_penalty": 1,
  • "early_stopping": false,
  • "stop_token_ids": [
    ],
  • "ignore_eos": false,
  • "min_tokens": 0,
  • "skip_special_tokens": true,
  • "spaces_between_special_tokens": true,
  • "truncate_prompt_tokens": 1,
  • "include_stop_str_in_output": false,
  • "response_format": {
    },
  • "guided_json": "string",
  • "guided_regex": "string",
  • "guided_choice": [
    ],
  • "guided_grammar": "string",
  • "guided_decoding_backend": "string",
  • "guided_whitespace_pattern": "string"
}

Response samples

Content type
application/json
null

Show Available Models

Responses

Request samples 

import requests

response = requests.get("http://example.com/v1/models")
print(response.json())

Response samples

Content type
application/json
null

Text Embeddings API

Compressa Embeddings API is OpenAI compatible API that allows you to create embeddings.

Create Embedding

Request Body schema: application/json
required
model
required
string (Model)
required
Array of Input (integers) or Array of Input (integers) or Input (string) or Array of Input (strings) (Input)
Encoding Format (string) or Encoding Format (null) (Encoding Format)
Default: "float"
Dimensions (integer) or Dimensions (null) (Dimensions)
User (string) or User (null) (User)
Additional Data (any) or Additional Data (null) (Additional Data)

Responses

Request samples

Content type
application/json
{
  • "model": "string",
  • "input": [
    ],
  • "encoding_format": "float",
  • "dimensions": 0,
  • "user": "string",
  • "additional_data": { }
}

Response samples

Content type
application/json
null

Rerank API

Compressa Rerank API allows you to rerank documents based on a query.

Rerank

Request Body schema: application/json
required
documents
required
Array of strings (Documents) [ 1 .. 2048 ] items [ items <= 122880 characters ]
query
required
string (Query) <= 122880 characters
return_documents
any (Return Documents)
Default: false
model
any (Model)
Default: "default/not-specified"

Responses

Request samples

Content type
application/json
{
  • "query": "string",
  • "documents": [
    ],
  • "return_documents": false,
  • "model": "default/not-specified"
}

Response samples

Content type
application/json
{
  • "object": "rerank",
  • "results": [
    ],
  • "model": null,
  • "usage": {
    },
  • "id": null,
  • "created": null
}

Layout API

Compressa Layout API allows you to partion documents.

Summary

Description

Request Body schema: multipart/form-data
required
files
required
string <binary>

The file to extract

xml_keep_tags
boolean (Xml Keep Tags)
Default: false

If True, will retain the XML tags in the output. Otherwise it will simply extract the text from within the tags. Only applies to partition_xml.

languages
Array of strings (OCR Languages)
Default: []

The languages present in the document, for use in partitioning and/or OCR

ocr_languages
Array of strings (OCR Languages)
Default: []

The languages present in the document, for use in partitioning and/or OCR

skip_infer_table_types
Array of strings (Skip Infer Table Types)
Default: []

The document types that you want to skip table extraction with. Default: []

Uncompressed Content Type (string) or Uncompressed Content Type (null) (Uncompressed Content Type)

If file is gzipped, use this content type after unzipping

output_format
string (Output Format)
Default: "application/json"
Enum: "application/json" "text/csv"

The format of the response. Supported formats are application/json and text/csv. Default: application/json.

coordinates
boolean (Coordinates)
Default: false

If true, return coordinates for each element. Default: false

Content type (string) or Content type (null) (Content type)

A hint about the content type to use (such as text/markdown), when there are problems processing a specific file. This value is a MIME type in the format type/subtype.

encoding
string (Encoding)
Default: "utf-8"

The encoding method used to decode the text input. Default: utf-8

Hi Res Model Name (string) or Hi Res Model Name (null) (Hi Res Model Name)

The name of the inference model used when strategy is hi_res

include_page_breaks
boolean (Include Page Breaks)
Default: false

If True, the output will include page breaks if the filetype supports it. Default: false

pdf_infer_table_structure
boolean (Pdf Infer Table Structure)
Default: true

Deprecated! Use skip_infer_table_types to opt out of table extraction for any file type. If False and strategy=hi_res, no Table Elements will be extracted from pdf files regardless of skip_infer_table_types contents.

strategy
string (Strategy)
Default: "auto"
Enum: "fast" "hi_res" "auto" "ocr_only"

The strategy to use for partitioning PDF/image. Options are fast, hi_res, auto. Default: auto

extract_image_block_types
Array of strings (Image block types to extract)
Default: []

The types of elements to extract, for use in extracting image blocks as base64 encoded data stored in metadata fields

unique_element_ids
boolean (unique_element_ids)
Default: false

When True, assign UUIDs to element IDs, which guarantees their uniqueness (useful when using them as primary keys in database). Otherwise a SHA-256 of element text is used. Default: False

"by_title" (string) or Chunking Strategy (null) (Chunking Strategy)

Use one of the supported strategies to chunk the returned elements. Currently supports: by_title

Combine Under N Chars (integer) or Combine Under N Chars (null) (Combine Under N Chars)

If chunking strategy is set, combine elements until a section reaches a length of n chars. Default: 500

max_characters
integer (Max Characters)
Default: 500

If chunking strategy is set, cut off new sections after reaching a length of n chars (hard max). Default: 1500

multipage_sections
boolean (Multipage Sections)
Default: true

If chunking strategy is set, determines if sections can span multiple sections. Default: true

New after n chars (integer) or New after n chars (null) (New after n chars)

If chunking strategy is set, cut off new sections after reaching a length of n chars (soft max). Default: 1500

overlap
integer (Overlap)
Default: 0

Specifies the length of a string ("tail") to be drawn from each chunk and prefixed to the next chunk as a context-preserving mechanism. By default, this only applies to split-chunks where an oversized element is divided into multiple chunks by text-splitting. Default: 0

overlap_all
boolean (Overlap all)
Default: false

When True, apply overlap between "normal" chunks formed from whole elements and not subject to text-splitting. Use this with caution as it entails a certain level of "pollution" of otherwise clean semantic chunk boundaries. Default: False

PDF Starting Page Number (integer) or PDF Starting Page Number (null) (PDF Starting Page Number)

When PDF is split into pages before sending it into the API, providing this information will allow the page number to be assigned correctly.

Responses

Request samples 

import requests

url = "https://compressa-api.mil-team.ru/v1/layout"
headers = {
    "Authorization": "Bearer TOKEN",
    "accept": "application/json",
}

files = {"files": open("path/to/file.pdf", "rb")}
data = {
    "xml_keep_tags": "false",
    "output_format": "application/json",
    "coordinates": "true",
    "strategy": "auto",
    "languages": ["rus", "eng"]
}

response = requests.post(
    url,
    headers=headers,
    files=files,
    data=data
)

print(response.json())

Response samples

Content type
application/json
[
  • {
    }
]