Improve AIMLAPI module defaults and chat overrides

Set Azure default base URL

Allow dict chat stream payloads

Default chat completions stream flag

Allow chat parse dict messages

Replaced all OpenAI imports and usage with AIMLAPI equivalents across example scripts, updated model names and endpoints, and refactored code to match AIMLAPI's API. Also removed obsolete enums and improved streaming/image handling logic. Deleted the unused responses_input_tokens.py example.

Handle streamed binary response in TTS payload

Adds a short-circuit to return StreamedBinaryAPIResponse directly in _finalize_sync_tts_payload, improving handling of streamed audio responses. Also adds a debug print statement in _coerce_mapping for error cases.

Handle AsyncStreamedBinaryAPIResponse in TTS payload

Adds a short-circuit to return the payload directly if it is an instance of AsyncStreamedBinaryAPIResponse in the _finalize_async_tts_payload function. This prevents unnecessary processing for streamed binary responses.

Add AIML audio STT/TTS overrides

Replaces all OpenAI imports and usage with AIMLAPI equivalents across example scripts. Updates model names, API key environment variables, and endpoint usage to match AIMLAPI conventions. Adjusts audio, Azure, and streaming examples for compatibility with AIMLAPI, and removes or comments out unsupported features.

Remove realtime audio example files

Deleted the examples/realtime directory including audio_util.py, azure_realtime.py, push_to_talk_app.py, and realtime.py. Updated background.py to use AIMLAPI instead of OpenAI. This streamlines the examples and updates the response client usage.

Introduce comprehensive tests for AIMLAPI client covering chat, completions, embeddings, error handling, image generation and editing, responses (basic, background, streaming), structured tool output, TTS, and chunked uploads. These tests use pytest and respx to mock HTTP interactions and validate request/response payloads and client behavior.

Audio translations are now explicitly unsupported in AIMLAPI, raising NotImplementedError with guidance for alternatives. Audio payload extraction logic in speech.py is enhanced to better handle various audio field formats and URLs, improving robustness and compatibility.

Update examples to use AIMLAPI client

Replaced OpenAI client imports and usage with AIMLAPI equivalents in async and streaming response examples. Also updated event type checks and removed a TODO comment for consistency.

Fix background JSON resume handling

Update image_stream.py

Update image_stream.py

Introduces AIMLAPI_HEADERS with referer and title for identification. These headers are merged into default_headers in _apply_default_client_options to ensure all requests include them.

Introduces image streaming and base64-encoded image response support to the AIMLAPI client. Adds Images and AsyncImages resources with b64_json hydration, patches response models for compatibility, and updates examples to demonstrate both URL and base64 image outputs. Removes the obsolete image_stream.py example.

Introduces the uploads resource to AIMLAPI and AzureAIMLAPI clients, including async variants. Implements Uploads and AsyncUploads classes to route upload requests to the files endpoint, ensuring compatibility with the API structure. Updates imports and __all__ in the uploads resource package for proper exposure.

Refactored comments in _version.py to clarify versioning scheme and added release automation marker. Cleaned up release-please-config.json by removing prerelease and extra file settings for improved configuration clarity.

Apply code style and formatting improvements

This commit updates code style for consistency, including quote normalization, line wrapping, and formatting in examples, source, and test files. It also updates pyproject.toml with new author information and adds 'ruff' as a dependency. No functional changes were made.

Update pyproject.toml

Updated repository references from 'api-sdk-python' to 'aimlapi-sdk-python' throughout the codebase and workflows. Changed version to 2.8.1-beta.1 in pyproject.toml, .release-please-manifest.json, and _version.py. Added a manual version bump workflow, removed OpenAI-specific release workflows, and adjusted CI and PyPI publishing logic for AIMLAPI. Cleaned up imports and ordering in several modules, improved README with example capabilities, and removed deprecated uploads example and test files.