Stream real-time responses from NanoChat, a 32-layer transformer language model built by Andrej Karpathy in 2025. Access this lightweight LLM via the NanoBrain platform for conversational AI tasks with streaming token-by-token output. | Input: Accepts an empty request body (no parameters required) to initiate a chat session | Output: Returns streaming server-sent events (SSE) with token-by-token text generation from the NanoChat d32 model, including model identification and completion markers | Use cases: I want to chat with a lightweight language model for quick responses; Get streaming AI responses for conversational tasks; I need to integrate a compact LLM into my application; Generate text responses using the NanoChat d32 model | Cost: 0.01 USDC on base

Extract the spreadsheet ID and worksheet ID (gid) from a Google Sheets URL.