Openai Chat

OpenAI Chat is a streaming chat client for any OpenAI-compatible Chat Completions endpoint. You supply your own API URL, API key, model name, system prompt, and temperature, then exchange messages with a model token-by-token using Server-Sent Events. The assistant response renders inline markdown (fenced code blocks, inline code, bold, italics), and both your conversation history and settings are persisted in the browser's localStorage so they survive page reloads.

1. Click Settings in the toolbar to open the configuration panel. 2. Fill in API URL (default https://api.openai.com/v1/chat/completions), API key (use the eye icon to reveal or hide it), Model (default gpt-4o-mini), System Prompt, and Temperature (0 to 2 slider). 3. Type a message in the input area and press Enter to send, or Shift+Enter for a newline. While the assistant is streaming, the Send button becomes a Stop button you can click to abort the request. 4. Use Clear chat in the toolbar to wipe the visible conversation and the stored history. The message counter at the right of the toolbar reflects how many user+assistant messages the current thread contains.

Requests are sent with stream:true and Authorization: Bearer <apiKey>, and the tool reads the SSE response chunk by chunk, appending each delta token to the assistant message in real time. Aborting a stream cleanly cancels the fetch via AbortController. Settings are saved under the localStorage key openai-chat-settings and the full message list under openai-chat-history, so the conversation is restored on next mount but lives only in your browser.