gen_ai_hub.proxy.native.google_genai package

Submodules

gen_ai_hub.proxy.native.google_genai.clients module

class AICoreDynamicTransport

Bases: BaseTransport

Synchronous transport that dynamically resolves deployment URLs per request.

__init__(proxy_client, **deployment_selector_kwargs)

Transport constructor.

Parameters:

proxy_client (BaseProxyClient) -- The proxy client used to select deployments.

close()

Closes the inner transport.

get_selector_kwargs()

Get the deployment selector kwargs.

Returns:

The deployment selector kwargs.

Return type:

dict

handle_request(request)

Handles the request by rewriting it to route through the appropriate AI Core deployment.

Parameters:

request (httpx.Request) -- The original HTTPX request.

Returns:

The HTTPX response from the AI Core deployment.

Return type:

httpx.Response

class AsyncAICoreDynamicTransport

Bases: AsyncBaseTransport

Asynchronous transport that dynamically resolves deployment URLs per request.

__init__(proxy_client, **deployment_selector_kwargs)

Transport constructor.

Parameters:

proxy_client (BaseProxyClient) -- The proxy client used to select deployments.

async aclose()

Closes the inner transport.

get_selector_kwargs()

get the deployment selector kwargs.

Returns:

The deployment selector kwargs.

Return type:

dict

async handle_async_request(request)

Handles the request by rewriting it to route through the appropriate AI Core deployment.

Parameters:

request (httpx.Request) -- The original HTTPX request.

Returns:

The HTTPX response from the AI Core deployment.

Return type:

httpx.Response

class Client

Bases: Client

Unified, native-feeling client that dynamically routes requests through SAP AI Core deployments based on the requested identifiers e.g. model name.

__init__(vertexai=True, project='placeholder', location='placeholder', deployment_id='', config_id='', config_name='', proxy_client=None, timeout=None, **kwargs)

Initializes the Client.

Parameters:

  • vertexai (bool, optional) -- to indicate Vertex AI usage, defaults to True

  • project (str, optional) -- the GCP project, defaults to "placeholder"

  • location (str, optional) -- the GCP location, defaults to "placeholder"

  • deployment_id (str, optional) -- the deployment identifier, defaults to ""

  • config_id (str, optional) -- the configuration identifier, defaults to ""

  • config_name (str, optional) -- the configuration name, defaults to ""

  • proxy_client (BaseProxyClient, optional) -- the proxy client to use, defaults to None

  • timeout (int, optional) -- the request timeout, defaults to None