gpt_api_free

This project is a large model API forwarding/proxy service; the repository mainly provides documentation and access guidance, not self-deployable software. Users obtain an API Key by registering and linking a GitHub account, then change the request address to the proxy Host provided by the project to call various mainstream large models. Main features (according to README): - Supports commonly used large models such as gpt, deepseek, claude, gemini, grok, qwen, kimi, minimax, glm. - The interface standard is consistent with the official OpenAI API, compatible with various software and plugins; the claude model also supports the official Anthropic protocol. - The free version offers a weekly quota of 50,000 points, a daily quota of 100 requests, and limits the call frequency to about 200 requests per day per IP & Key. - Provides a domestic proxy Host (api.chatanywhere.tech) and an overseas Host (api.chatanywhere.org); in the domestic environment no proxy is needed to use. - The paid version bills based on the token usage returned by the official providers, balance does not expire, and the README claims the price is lower than the official rates. Usage: Register and bind GitHub to obtain a Key, replace the request address with the proxy Host, and change the model name as needed. The README provides tutorial links for common software and a status page to facilitate troubleshooting service availability issues. Notes: Free Keys are limited to personal non-commercial use, education, and non-profit research; commercial use and large-scale training are prohibited. The README states that user input text is not collected, but upstream providers may retain data according to their data policies. Prices and model lists may change; refer to the official website for the latest information.