[Feature Request] Use the OpenAI Responses API for the openai-like backend (follow-up to #13440)

Description #13440 was closed as an upstream llama-index gap. I tested this on 3.2.1 and the diagnosis is correct, but the problem is recurring: models llama-index already lists (gpt-5.4-mini, gpt-5.5) work over /v1/chat/completions, while every newer model it doesn't list yet (gpt-5.6-luna, gpt-6-luna) fails AI suggestions with: 400 Function tools with reasoning_effort are not supported for <model> in /v1/chat/completions. To use function tools, use /v1/responses or set reasoning_effort to 'none'. So each new OpenAI model is unusable until both a llama-index release and a Paperless dependency bump land. PAPERLESS_AI_LLM_EXTRA_PARAMS={"reasoning_effort":"none"} avoids the error, but only by switching reasoning off. Proposal: an opt-in PAPERLESS_AI_LLM_USE_RESPONSES_API=true that makes the openai-like backend use llama-index's existing OpenAIResponses client (/v1/responses). No model-specific code: the configured llm_context_size …