OpenAI compatible interface physical examination CLI, one command to check the model list of any transfer station, self-built gateway or local inference service, SSE streaming and usage blocks, tool calls, JSON mode, Responses API (to determine whether it can be connected to Codex), image input, error format and first word delay, output desensitized JSON / Markdown report; single file, only relies on the Python standard library, and can also be used as a GitHub Action Run regularly; only checks protocol compatibility, does not prove the authenticity of the model, MIT open source
Description translated from the developer's original Chinese by machine translation. 中文原页 ↗
Free online toolbox, a collection of 32 ready-to-use daily gadgets (image compression and cropping, JPG/PNG/WebP/HEIC format conversion, color selection and conversion, QR code and password generation, word count, BMI/tip/discount calculator, etc.), all run locally in the browser, no registration is required, no files are uploaded, and the interface supports 7 languages
Online playground and access guide for the Mistral Large 4 model, 1M context, supports text and image input, adjustable inference gear, temperature, maximum output token and JSON output format; the playground is free, and points are charged for exceeding the free limit (points are the US dollar cost of the model × 12,000), non-Mistral official site
Image generation and editing workbench, focusing on Nano Banana 2.1 and 2.1 Pro (Pro is the 7-point 4K preset of the same model), optional 1K/2K/4K, the editing mode supports uploading 1 to 14 reference images; the bottom layer runs the Google Gemini image model, an unofficial site, new accounts will be given 3 points at a time, and the billing is based on points
Mac disk cleaning tool, check the usage of WeChat pictures and videos, find duplicate files, preview and select items one by one before cleaning; scanning and viewing results are free, the full function can be tried for 7 days, cleaning after the trial requires a license, supports macOS 14+
Upload a photo of a cat or dog, choose one of the 16 Halloween looks, and it will take about 30 seconds to generate a high-definition cross-dressing photo that retains the breed, coat color, and facial features; free watermark preview without registration. After registration, you can save the high-definition picture. The high-definition package starts at $1.99 for one-time payment without subscription.
Upload one to three photos to generate a trendy short video of a gas station before sunset. The three shots of dancing, selfies, and getting off the car are replaced with the faces of you and your friends one by one. The scene and camera movements are fixed. Output is a 10-second 9:16 silent MP4; billed by points
About this listing
This page is part of Indie Maker, a directory of works by Chinese indie developers. We sync daily from the open-source repository chinese-independent-developer; this product was added on 2026-10-11.
About pricingAny pricing, membership, plans or premium features mentioned in a product's description belong to that product itself, not to this directory. We don't build, host or resell any product — the directory is entirely free.