Implementation:Cohere ai Cohere python BatchesRawClient

Knowledge Sources	Cohere Python SDK
Domains	SDK, API_Client
Last Updated	2026-02-15 14:00 GMT

Overview

The RawBatchesClient and AsyncRawBatchesClient classes provide the low-level HTTP client for batch operations against the Cohere v2/batches API endpoint, returning typed HttpResponse wrappers that include both the parsed data and raw HTTP response metadata.

Description

This file is auto-generated by Fern from the Cohere API definition. It contains two classes:

• RawBatchesClient -- The synchronous low-level client that directly issues HTTP requests to the Cohere batch API. Each method constructs the HTTP request using the underlying SyncClientWrapper, sends it via httpx, parses the JSON response into typed models using construct_type, and returns an HttpResponse[T] wrapper. Error responses are mapped to specific exception types based on HTTP status code.
• AsyncRawBatchesClient -- The asynchronous counterpart that wraps AsyncClientWrapper and provides the same methods as async coroutines, returning AsyncHttpResponse[T].

The raw client handles the following HTTP error mappings:

• 400 -- BadRequestError
• 401 -- UnauthorizedError
• 403 -- ForbiddenError
• 404 -- NotFoundError
• 500 -- InternalServerError
• 503 -- ServiceUnavailableError

For any other non-2xx status code, a generic ApiError is raised.

Usage

The raw client is typically not used directly by end users. Instead, it is wrapped by BatchesClient which extracts the .data attribute for convenience. Use the raw client directly when you need access to HTTP response metadata such as status codes and headers, accessible via client.batches.with_raw_response.

Code Reference

Source Location

• Repository: Cohere Python SDK
• File: src/cohere/batches/raw_client.py (926 lines)

Signature

class RawBatchesClient:
    def __init__(self, *, client_wrapper: SyncClientWrapper): ...

    def list(
        self,
        *,
        page_size: typing.Optional[int] = None,
        page_token: typing.Optional[str] = None,
        order_by: typing.Optional[str] = None,
        request_options: typing.Optional[RequestOptions] = None,
    ) -> HttpResponse[ListBatchesResponse]: ...

    def create(
        self,
        *,
        request: Batch,
        request_options: typing.Optional[RequestOptions] = None,
    ) -> HttpResponse[CreateBatchResponse]: ...

    def retrieve(
        self,
        id: str,
        *,
        request_options: typing.Optional[RequestOptions] = None,
    ) -> HttpResponse[GetBatchResponse]: ...

    def cancel(
        self,
        id: str,
        *,
        request_options: typing.Optional[RequestOptions] = None,
    ) -> HttpResponse[CancelBatchResponse]: ...


class AsyncRawBatchesClient:
    def __init__(self, *, client_wrapper: AsyncClientWrapper): ...

    async def list(...) -> AsyncHttpResponse[ListBatchesResponse]: ...
    async def create(...) -> AsyncHttpResponse[CreateBatchResponse]: ...
    async def retrieve(...) -> AsyncHttpResponse[GetBatchResponse]: ...
    async def cancel(...) -> AsyncHttpResponse[CancelBatchResponse]: ...

Import

from cohere.batches.raw_client import RawBatchesClient, AsyncRawBatchesClient

I/O Contract

Inputs

list()

Name	Type	Required	Description
page_size	typing.Optional[int]	No	Maximum number of batches to return. Default is 50; maximum is 1000. Values above 1000 are coerced to 1000.
page_token	typing.Optional[str]	No	A page token received from a previous ListBatches call. Used for pagination.
order_by	typing.Optional[str]	No	Ordering field. Use created_at for creation time or updated_at for last updated time.
request_options	typing.Optional[RequestOptions]	No	Request-specific configuration (timeouts, headers, etc.).

create()

Name	Type	Required	Description
request	Batch	Yes	The batch definition object containing name, input_dataset_id, and model. Serialized to JSON via convert_and_respect_annotation_metadata.
request_options	typing.Optional[RequestOptions]	No	Request-specific configuration.

retrieve()

Name	Type	Required	Description
id	str	Yes	The batch ID. Encoded into the URL path as v2/batches/{id}.
request_options	typing.Optional[RequestOptions]	No	Request-specific configuration.

cancel()

Name	Type	Required	Description
id	str	Yes	The batch ID. Encoded into the URL path as v2/batches/{id}/cancel.
request_options	typing.Optional[RequestOptions]	No	Request-specific configuration.

Outputs

Method	Type	Description
list()	HttpResponse[ListBatchesResponse]	HTTP response wrapper containing parsed ListBatchesResponse data and raw response metadata.
create()	HttpResponse[CreateBatchResponse]	HTTP response wrapper containing parsed CreateBatchResponse data and raw response metadata.
retrieve()	HttpResponse[GetBatchResponse]	HTTP response wrapper containing parsed GetBatchResponse data and raw response metadata.
cancel()	HttpResponse[CancelBatchResponse]	HTTP response wrapper containing parsed CancelBatchResponse data and raw response metadata.

HTTP Endpoints

Method	HTTP Verb	Endpoint
list()	GET	v2/batches
create()	POST	v2/batches
retrieve()	GET	v2/batches/{id}
cancel()	POST	v2/batches/{id}/cancel

Error Handling

HTTP Status	Exception Class
400	BadRequestError
401	UnauthorizedError
403	ForbiddenError
404	NotFoundError
500	InternalServerError
503	ServiceUnavailableError
Other non-2xx	ApiError

Usage Examples

from cohere import Client

client = Client(
    client_name="YOUR_CLIENT_NAME",
    token="YOUR_TOKEN",
)

# Access the raw client via the with_raw_response property
raw_response = client.batches.with_raw_response.list(page_size=10)

# Access HTTP metadata
print(raw_response.status_code)
print(raw_response.headers)

# Access the parsed data
batches_data = raw_response.data
print(batches_data)

Related Pages

• Environment:Cohere_ai_Cohere_python_Python_SDK_Runtime