- **ApiAirforce.py**: Added `use_image_size = True` class attribute.
- **EasyChat.py**: Added `user_data_dir=None` argument when calling `get_args_from_nodriver`.
- **LMArenaBeta.py**:
- Only set `cls.share_url` from `G4F_SHARE_URL` if `cls.share_url` is `None`.
- Added `user_data_dir=None` argument when calling `get_args_from_nodriver`.
- Modified media filtering to check `url` is a `str` before `startswith("https://")`.
- **OpenaiTemplate.py**:
- Added `use_image_size = False` class attribute.
- Modified image generation payload to optionally include `"size"` key when `use_image_size` is `True` and width/height present.
- Refactored `read_response` to store `content` in a variable, strip on first chunk, and yield `ToolCalls` directly from `tool_calls` variable.
- **g4f/requests/__init__.py**:
- Added `user_data_dir` parameter (default `"nodriver"`) to `get_args_from_nodriver`.
- Passed `user_data_dir` to `get_nodriver` call.
- g4f/Provider/ApiAirforce.py: replace class attr parent="Api.Airforce" with label="Api.Airforce"
- g4f/Provider/EasyChat.py: add guard after loop to raise ValueError("Failed to retrieve arguments for EasyChat.") when args is falsy
- g4f/Provider/SearXNG.py:
- add imports: from .helper import format_media_prompt; from .. import debug
- rename class attr default_url to url (SEARXNG_URL) and use cls.url in requests
- create(): add parameters prompt: str = None and language: str = "it"
- build prompt via format_media_prompt(messages, prompt); remove previous query extraction
- request params: set q to prompt and language to provided language
- pass proxy to session.get(...); replace print(...) with debug.log(...)
- on empty results, return early (remove "Nessun risultato trovato." and FinishReason("stop") yields)
- result formatting: use r.get("title") without default fallback
- g4f/Provider/__init__.py: import DuckDuckGo and Free2GPT from .deprecated.DuckDuckGo and .deprecated.Free2GPT
- g4f/Provider/deprecated/DuckDuckGo.py: update relative imports to ...typing, ..base_provider, ..helper
- g4f/Provider/deprecated/Free2GPT.py: update relative imports to ...errors, ...requests, ...typing, ..base_provider
- g4f/Provider/needs_auth/DeepSeekAPI.py: set working = False (was has_dsk)
- g4f/Provider/needs_auth/You.py: set working = False (was True); update headers["Referer"] to f"{cls.url}/api/streamingSearch"
- **g4f/Provider/Copilot.py**
- Added `"Smart (GPT-5)"` to `models` list.
- Added `"gpt-5"` alias mapping to `"GPT-5"` in `model_aliases`.
- Introduced `mode` selection logic to support `"smart"` mode for GPT-5 models alongside existing `"reasoning"` and `"chat"` modes.
- **g4f/Provider/EasyChat.py**
- Added `get_models` class method to map `-free` models to aliases and store them in `cls.models`.
- Resolved model via `cls.get_model(model)` at start of `create_async_generator`.
- Reset `cls.captchaToken` to `None` at the beginning of `callback`.
- Wrapped main generator logic in a loop to allow retry once if `CLEAR-CAPTCHA-TOKEN` error occurs, clearing auth file and resetting args.
- **g4f/Provider/needs_auth/OpenaiChat.py**
- Added handling for image models: detect and set `image_model` flag, use `default_model` when sending requests if image model selected, and include `"picture_v2"` in `system_hints` when applicable.
- Replaced textarea/button detection code in page load sequence with `nodriver` `select` calls, sending "Hello" before clicking send button, and included profile button selection if class needs auth.
- **g4f/Provider/openai/models.py**
- Changed `default_image_model` from `"dall-e-3"` to `"gpt-image"`.
- Added `"gpt-5"` and `"gpt-5-thinking"` to `text_models` list.
- Added alias mapping for `"dall-e-3"` pointing to new `default_image_model`.
- Added `await asyncio.sleep(1)` inside captcha verification loop in `EasyChat.py` to introduce delay between checks
- Modified `Grok.py` to send "Hello" input to a selected textarea element during auth flow
- Added delay after sending keys to textarea in `Grok.py` using `await asyncio.sleep(1)`
- Added logic to select and click a submit button if present in `Grok.py` during header check loop
- All changes are within the `EasyChat` and `Grok` class definitions respectively
- Added new `EasyChat` provider (`g4f/Provider/EasyChat.py`) with captcha handling, nodriver callback, and token caching
- Added new `GLM` provider (`g4f/Provider/GLM.py`) with model retrieval, auth token fetch, and SSE streaming support
- Updated `g4f/Provider/__init__.py` to import `EasyChat` and `GLM`
- Modified `LMArenaBeta` in `g4f/Provider/needs_auth/LMArenaBeta.py` to remove nodriver availability check and always use `get_args_from_nodriver` with callback
- Updated `HuggingFaceAPI` in `g4f/Provider/needs_auth/hf/HuggingFaceAPI.py` to use `default_model` from `models` instead of `default_llama_model` and removed commented `max_inputs_lenght` param
- Updated `HuggingFace` in `g4f/Provider/needs_auth/hf/__init__.py` to import `default_model` instead of `default_vision_model`, set `default_model` class attribute, and commented out HuggingFaceInference and image model handling logic
- Modified `OpenaiTemplate` in `g4f/Provider/template/OpenaiTemplate.py` to prefer `"name"` over `"id"` when populating `vision_models`, `models`, and `models_count`
- Enhanced `sse_stream` in `g4f/requests/__init__.py` to strip and skip empty `data:` lines, handle JSON decode errors, and raise `ValueError` on invalid JSON
