0xmd

0xmd is a healthcare generative AI from Hong Kong's 0X LIMITED. Its public face is a free, signup-free multilingual health Q&A app and website; underneath sit clinical teaching modules for clinics and enterprises plus an OpenAI-compatible API.

Core capabilities

• Health Q&A: ask by text or voice; it analyzes first, then routes, returning a multi-step reasoning chain, with support for 50+ languages
• Scan interpretation: upload checkup reports, medicine boxes and food labels, skin photos, X-rays and MRIs to get structured explanations
• Wearable sync: connects to Apple Health and Health Connect, folding activity, sleep, heart rate and body fat into "My Week"
• Nooduo module: gamified health-knowledge challenges, with 12 learning paths that include USMLE and NCLEX question banks
• Clinical teaching modules: 16 modules covering clinic coding and claims, dentistry, prehospital emergency care, tuberculosis screening and virtual patients
• Developer API: POST /chat/completions, with question_type options xray, skin_disease and wound

Who it suits Everyday users who want to read their checkup reports, medicine boxes and imaging; developers embedding a health Q&A entry point into their site; observers watching the Hong Kong and Brazil markets. People who need diagnostic conclusions, medication decisions or an auditable model should not use it.

Distinctive strengths It combines traditional Chinese and Western medicine, multiple languages and multimodal handling of reports and imaging, leaving open ground in Chinese-speaking and Portuguese-speaking markets; comparable free products stop at plain chat, while this one wires up scanning and wearables as well.

Editorial assessment It is fairly complete, and its compliance text is more careful than at startups of the same stage, but a few things need saying: the 0xmd MoE-LLMs claim appears only in the company's own description, with no model card, no paper, no weights and no leaderboard entry, and the only score is an internal 89.0 MedQA benchmark on the showcase page, while the HKTDC profile gives 80.15 on USMLE — the figures conflict, and as things stand there is no verifiable third-party evaluation. The terms deny qualification in all four jurisdictions, FDA, NMPA, MDACS and ANVISA, state that no licensed physician reviews AI answers, and cap liability at 100 US dollars because the service is free; the mainland section says outright that the service is not offered in mainland China. There are contradictions inside the site too: the cookie policy says there is no ad network, yet the homepage loads ad scripts and the stores mark the app as containing ads. G2, Capterra and Trustpilot have zero entries, iOS has zero ratings, and Play shows only 10+ downloads.