LLM Puzzles vs SSE Inspector

Word games powered by a local LLMs. Investigating the limits of on-device LLMs. Is a model small enough to fit on a device smart enough to be useful? Download and store multiple models, and have them play games and chat to evaluate their usefulness.

SSE Inspector is a focused stream debugger for developers building with AI APIs and LLMs. It preserves the original SSE (Server-Sent Events) response and presents every packet, semantic delta, packet interval, response header, and assembled output so broken, stalled, merged, or incomplete streams are easier to diagnose. It supports OpenAI, DeepSeek, Claude, Gemini, Ollama, Llama, and custom SSE mappings. Paste a cURL command to instantly import its URL, method, headers, body, and credentials; the app automatically detects the protocol from endpoint and request characteristics. Sessions can be edited, rerun, replayed, imported, exported, and shared as redacted Debug Bundles. No account required. Sessions and credentials remain on device and are sent only to endpoints you explicitly run. Initial Release! A powerful companion for AI & LLM developers to debug Server-Sent Events. AI SSE Presets: Ready-to-use environments for OpenAI, DeepSeek, Claude, Gemini, and Ollama/Llama. Smart cURL Import: Paste cURL to automatically detect and map protocols. Deep Diagnostics: Raw SSE stream viewing, real-time response rendering, timing charts, and interval analytics. Session Management: Edit, rerun, replay, import, and export debugging sessions with ease.