AIMessageChunk(content='') for thinking tokens missng from vllm instance

feat(partners): ContextOverflowError for Groq/Mistral + standard test coverage

Missing model field huggingface_endpoint.py

core: default_tool_parser crashes on malformed function dicts, silently drops all valid tool calls

To support BytePlus's video functionality, implement lenient validation for the input.type field in OpenAI responses by checking if input.type == input_video to ensure compatibility with BytePlus.

AIMessageChunk(content='') for thinking tokens missng from vllm instance

feat(partners): ContextOverflowError for Groq/Mistral + standard test coverage

Missing model field huggingface_endpoint.py

core: default_tool_parser crashes on malformed function dicts, silently drops all valid tool calls

To support BytePlus's video functionality, implement lenient validation for the input.type field in OpenAI responses by checking if input.type == input_video to ensure compatibility with BytePlus.