purra-openai · Python

English | 简体中文

OpenAI model gateways for PurrA using the official Python SDK. Supports text, function tools, streaming, cancellation, and token usage. Requires Python 3.11+.

In 1.0.0 development, Chat Completions also accepts the optional static image contract with image_input=True. Responses image input is not implemented in this slice.

Install

From the repository root:

python -m pip install . ./integrations/openai/python

Choose an API

from openai import AsyncOpenAI
from purra_openai import OpenAIResponsesGateway

client = AsyncOpenAI()
gateway = OpenAIResponsesGateway(client, timeout_seconds=60)

For Chat Completions, replace the gateway:

from purra_openai import OpenAIChatCompletionsGateway

gateway = OpenAIChatCompletionsGateway(client, timeout_seconds=60)

Pass gateway to AgentCore(model_gateway=gateway, ...). In the Run's ModelRequest, set provider="openai", the model ID, and its capability snapshot. The application supplies credentials and verified model limits.

Options and behavior

ModelRequest.options accepts max_tokens, temperature, top_p, and reasoning_effort. Choose options supported by the selected model. Core resolves the output limit; the gateway applies it to the selected API.

SDK retries are disabled so Core controls retry and budget accounting. Server-side response storage is disabled. Responses reasoning continuation is kept in private message data; preserve it with the conversation when resuming.

These gateways support native OpenAI text and function-tool APIs. Images, audio, and hosted tools are not supported. Other vendors require application adapters.

Close a supplied SDK client in the application. await gateway.close() closes only a client created by the gateway itself.