genai package

class genai.ApiClient[source]

Bases: object

Class which provides methods for making API requests. It provides methods for obtaining synchronous and asynchronous HTTP clients.

Config

alias of BaseConfig

__init__(*, credentials, config=None)[source]
Parameters:

  • credentials (Credentials)

  • config (dict | BaseConfig | None)

get_async_http_client(rate_limit_options=None, retry_options=None, client_options=None, **kwargs)[source]

Gets cached instance of AsyncHttpxClient (cache key is created from provided parameters).

Raises:

  • ValueError – If both rate_limit_options and retry_options are provided.

  • RuntimeError – If the method is not called within an asynchronous environment.

Parameters:

  • rate_limit_options (dict | None)

  • retry_options (dict | None)

  • client_options (dict | None)

Return type:

AsyncHttpxClient

get_http_client(retry_options=None, **kwargs)[source]
Parameters:

retry_options (dict | None)

Return type:

HttpxClient

class genai.Client[source]

Bases: BaseService[BaseConfig, BaseServices]

The Client class provides an interface for interacting with various services through an API client. It can be initialized with either an api_client or credentials along with optional configurations and services.

Example:

from genai import Credentials, Client

credentials = Credentials.from_env()
client = Client(credentials=credentials)
text

An instance of the TextService class for text-related operations.

request

An instance of the RequestService class for making request-related operations.

tune

An instance of the TuneService class for tuning models.

model

An instance of the ModelService class for managing models.

file

An instance of the FileService class for managing files.

prompt

An instance of the PromptService class for working with prompts.

system_prompt

An instance of the SystemPromptService class for working with system prompts.

user

An instance of the UserService class for managing user-related operations.

tag

An instance of the TagService class for working with tags.

folder

An instance of the FolderService class for working with folder.

task

An instance of the TaskService class for working with tasks.

Config

alias of BaseConfig

Services

alias of BaseServices

__init__(*, api_client: ApiClient, config: dict | BaseConfig | None = None, services: BaseServices | None = None) → None[source]
__init__(*, credentials: Credentials, config: dict | BaseConfig | None = None, services: BaseServices | None = None) → None
Parameters:

  • credentials – The credentials used to authenticate the API client.

  • api_client – The API client used to make requests to the API.

  • config – The configuration for the API client.

  • services – The services object containing instances of various service classes.

Raises:

ValueError – Either ‘api_client’ or ‘credentials’ needs to be passed.

Note

The api_client parameter must be provided either directly or through credentials. If no services parameter is provided, the client will use the default one.

pydantic model genai.Credentials[source]

Bases: BaseModel

The Credentials class represents credentials required for accessing the GENAI API.

- `api_key`

API key which can be retrieved from the UI.

- `api_endpoint`

A string representing the GENAI API endpoint (default is BAM).

Type:

optional

Examples

Create a Credentials instance with explicit api_endpoint:

credentials = Credentials(api_key="your_api_key", api_endpoint="https://bam-api.res.ibm.com")

Create a Credentials instance with default api_endpoint:

credentials = Credentials(api_key="your_api_key")

Create a Credentials instance from environment variables:

credentials = Credentials.from_env()
Validators:
field api_endpoint: str [Required]

GENAI API Endpoint

Constraints:

  • min_length = 1

Validated by:

  • _validate_api_endpoint

field api_key: SecretStr [Required]

The GENAI API Key

__init__(api_key, api_endpoint=None, **kwargs)[source]

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

  • api_key (str)

  • api_endpoint (str | None)

classmethod from_env(*, api_key_name=None, api_endpoint_name=None)[source]
Parameters:

  • api_key_name (str | None)

  • api_endpoint_name (str | None)

Return type:

Credentials

genai.handle_shutdown_event(*_args)[source]

Handles the shutdown event.

Return type:

None

Subpackages

Submodules