JinzeAI
One OpenAI-compatible API for multiple AI models
Start new thread

JinzeAI - One OpenAI-compatible API for multiple AI models

by•
Use your existing OpenAI SDK with supported DeepSeek, Qwen, Kimi, GLM, and MiniMax models through one endpoint. JinzeAI is in public beta for users outside mainland China, with limited free test credit and no live payments during beta.

Add a comment

Replies

Best
Maker
📌
Hi Product Hunt — I’m Jinze, an independent maker building JinzeAI. I started this project because testing several AI model providers often means managing different SDKs, endpoints, keys, and error formats. JinzeAI provides one OpenAI-compatible endpoint so developers can keep a familiar client and change the model name when testing supported models. The public beta currently focuses on three things: 1. Clear OpenAI-compatible setup for Python, Node.js, and cURL. 2. A small set of supported models from DeepSeek, Qwen, Kimi, GLM, and MiniMax. 3. Practical feedback on SDK compatibility, latency, and error messages. There are no live payments during the public beta. New testers may receive limited free test credit. Model availability and limits can change while the service is being validated. I would especially appreciate feedback from developers who already have an OpenAI-compatible application and can test how easily it switches over. Website: Documentation: Examples: