MediBot / README.md
github-actions[bot]
Deploy MediBot from ac379a12
4dc72d3
metadata
title: 'MediBot: Free AI Medical Assistant · 20 languages'
emoji: 🏥
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860
pinned: true
license: apache-2.0
short_description: Free AI medical chatbot. 20 languages. No sign-up.
tags:
  - medical
  - healthcare
  - chatbot
  - medical-ai
  - health-assistant
  - symptom-checker
  - telemedicine
  - who-guidelines
  - cdc
  - multilingual
  - i18n
  - rag
  - llama-3.3
  - llama-3.3-70b
  - mixtral
  - groq
  - huggingface-inference
  - pwa
  - offline-first
  - free
  - no-signup
  - privacy-first
  - worldwide
  - nextjs
  - docker
models:
  - meta-llama/Llama-3.3-70B-Instruct
  - meta-llama/Meta-Llama-3-8B-Instruct
  - mistralai/Mixtral-8x7B-Instruct-v0.1
  - Qwen/Qwen2.5-72B-Instruct
  - deepseek-ai/DeepSeek-V3
  - ruslanmv/Medical-Llama3-8B
  - google/gemma-2-9b-it
datasets:
  - ruslanmv/ai-medical-chatbot

MediBot — free AI medical assistant, worldwide

Tell MediBot what's bothering you. In your language. Instantly. For free. No sign-up. No paywall. No data retention. Aligned with WHO · CDC · NHS guidelines.

Try MediBot Languages Free No sign-up

Why MediBot

  • Free forever. No API key, no sign-up, no paywall, no ads.
  • 20 languages, auto-detected. English, Español, Français, Português, Deutsch, Italiano, العربية, हिन्दी, Kiswahili, 中文, 日本語, 한국어, Русский, Türkçe, Tiếng Việt, ไทย, বাংলা, اردو, Polski, Nederlands.
  • Worldwide. IP-based country detection picks your local emergency number (190+ countries) and adapts the answer to your region (°C/°F, metric/imperial, local guidance).
  • Best free LLM on HuggingFace. Powered by Llama 3.3 70B via HF Inference Providers (Groq) — fastest high-quality free tier available — with an automatic fallback chain across Cerebras, SambaNova, Together, and Mixtral.
  • Grounded on WHO, CDC, NHS, NIH, ICD-11, BNF, EMA. A structured system prompt aligns every answer with authoritative guidance.
  • Red-flag triage. Built-in symptom patterns detect cardiac, neurological, respiratory, obstetric, pediatric, and mental-health emergencies in every supported language — and immediately escalate to the local emergency number.
  • Installable PWA. Add to your phone's home screen and use it like a native app. Offline-capable with a cached FAQ fallback.
  • Shareable. Every AI answer gets a Share button that generates a clean deep link with a branded OG card preview — perfect for WhatsApp, Twitter, and Telegram.
  • Private & anonymous. Zero accounts. Zero server-side conversation storage. No IPs logged. Anonymous session counter only.
  • Open source. Fully transparent. github.com/ruslanmv/ai-medical-chatbot

How it works

  1. You type (or speak) a health question
  2. MedOS checks for emergency red flags first
  3. It searches a medical knowledge base for relevant context
  4. Your question + context go to Llama 3.3 70B (via Groq, free)
  5. You get a structured answer: Summary, Possible causes, Self-care, When to see a doctor

If the main model is busy, MedOS automatically tries other free models until one responds.

Built with

Layer Technology
Frontend Next.js 14, React, Tailwind CSS
AI Model Llama 3.3 70B Instruct (via HuggingFace Inference + Groq)
Fallbacks Mixtral 8x7B, OllaBridge, cached FAQ
Knowledge Medical RAG from ruslanmv/ai-medical-chatbot dataset
Gateway OllaBridge-Cloud
Hosting HuggingFace Spaces (Docker)

License

Apache 2.0 — free to use, modify, and distribute.