QubitPage® OS desktop — A full web-based quantum OS with IBM Quantum integration, AI drug discovery, real medical research tools, and a MedGemma disease diagnosis assistant running in the browser. Fully responsive across desktop, tablet, and mobile — with draggable desktop icons and touch support.
| Repository | Description | Status |
|---|---|---|
| QubitPage-OS | ← This repo — Full Quantum OS Platform | ✅ Live |
| QuBIOS | Transit Ring quantum middleware engine | ✅ Live |
| QLang | Quantum Programming Language + Browser SDK | ✅ Live |
QubitPage® OS is a complete browser-based quantum operating system that provides:
- Desktop environment — A full windowed OS in the browser with taskbar, app launcher, multi-window support, draggable desktop icons (mouse & touch), and fully responsive layout (mobile, tablet, desktop)
- Quantum Circuit Lab — Write and run QLang circuits on real IBM Quantum hardware or local simulators
- AI Drug Discovery — Quantum-enhanced molecular simulation for diseases without cures (GBM, TB, Alzheimer's, ALS, IPF)
- MedGemma AI — Google's medical AI for disease diagnosis, ADMET prediction, and treatment analysis
- QuBIOS Transit Ring — 5× qubit lifespan extension with 99.80% Bell state fidelity
- ARIA AI Assistant — Gemini-powered research assistant integrated into every tool
- Real Research Results — 13 novel drug candidates discovered, IBM Fez real hardware validation
┌─────────────────────────────────────────────────────────────────────┐
│ QubitPage® OS v1.1.0 │
│ Browser Desktop (os.html) │
├──────────┬──────────┬──────────┬──────────┬──────────┬─────────────┤
│ Circuit │ Quantum │ ARIA AI │MedGemma │ QuBIOS │ QuantumTB │
│ Lab │ Drug │Assistant │ Diag. │ QubiLgc │ QuantumNrο │
│ (QLang) │ Sim │ (Gemini) │ Port5051 │TransitRg │ Disease Hub│
├──────────┴──────────┴──────────┴──────────┴──────────┴─────────────┤
│ quantum_kernel.py │
│ quantum_backends.py (IBM / AWS / Simulators) │
├─────────────────────────────────────────────────────────────────────┤
│ qubilogic.py — QuBIOS Transit Ring Engine │
│ TransitRing | SteaneQEC | TeleportEngine | EntanglementDistiller │
├─────────────────────────────────────────────────────────────────────┤
│ IBM Quantum / Stim / Cirq / Qiskit │
└─────────────────────────────────────────────────────────────────────┘
git clone https://github.com/qubitpage/QubitPage-OS.git
cd QubitPage-OS
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
# Set API keys (see INSTALL.md for full guide)
export GEMINI_API_KEY=your_key_here
export IBM_QUANTUM_TOKEN=your_token_here # optional, simulators work without it
export GROQ_API_KEY=your_key_here
python3 src/app.py
# → Open http://localhost:5050Full setup guide: INSTALL.md
| App | Icon | Description |
|---|---|---|
| Terminal | >_ |
QLang quantum shell interpreter |
| Circuit Lab | ⚛ | Visual quantum circuit builder + IBM/Stim runner |
| Quantum Oracle | 🎮 | Interactive quantum algorithm playground |
| Crypto Tools | 🔐 | Quantum encryption, superdense coding, BB84 |
| ARIA Assistant | 🤖 | Gemini-powered AI research assistant |
| QuBIOS / QubiLogic | 🧠 | QuBIOS Transit Ring memory + entanglement lab |
| QuantumNeuro | 🧠 | GBM (glioblastoma) quantum drug discovery |
| QuantumTB | 🫁 | Tuberculosis DprE1 inhibitor research |
| Disease Hub | 🏥 | Multi-disease quantum drug screening dashboard |
| MedGemma AI | 🏥 | AI-assisted medical diagnosis + treatment search |
| MedLab | 🗂 | Real medical case analysis engine |
| Quantum Search | 🔍 | Grover's algorithm drug target search |
| Quantum Drug | 🧬 | Molecular quantum simulation & ADMET scoring |
| Training Results | 🎯 | View all drug screening & research training runs |
| Discovery Reports | 📊 | 13 novel drug candidates with QBP-### IDs |
| Med Files | 📁 | Patient data + lab report processing |
| Docs | 📚 | Full platform documentation |
| QuantumDerm | 🔬 | AI-powered skin lesion analysis + drug screening (MedGemma 4B + Derm Foundation) |
| Settings | ⚙ | User preferences, API keys, backend config |
- MedGemma 4B — Google's medical-domain instruction-tuned model for diagnosis reasoning, disease classification, and ADMET prediction
- Derm Foundation — EfficientNet-B0 dermatology classifier (99.73% val accuracy) used by QuantumDerm
- CXR Foundation — Chest X-ray DenseNet-121 (100% val accuracy)
- Path Foundation — ViT-B/16 pathology foundation model (100% val accuracy)
- Gemini 2.0 Flash — Powers ARIA assistant for research synthesis and drug target analysis
- TxGemma — Therapeutic property prediction for drug screening
- IBM Quantum (ibm_fez, ibm_sherbrooke) — Real 127–156 qubit hardware via Qiskit
- Stim — Ultra-fast local Clifford circuit + error correction simulator
- Qiskit Statevector — Exact local simulation (≤32 qubits)
- Google Cirq (docs: see below) — Via Cirq local + Google Quantum Computing Service
See docs/ai-models.md for full model guide.
See docs/quantum-simulators.md for all backends.
| ID | Target | Disease | Predicted Activity |
|---|---|---|---|
| QBP-007 | rpoB | Tuberculosis | BBB-penetrant, MIC↓ |
| QBP-001 | KRAS G12D | GBM | Novel covalent scaffold |
| QBP-004 | Aβ42 | Alzheimer's | Aggregation inhibitor |
| QBP-006 | MUC5B | IPF (lung fibrosis) | Anti-fibrotic |
| QBP-005 | TGF-β ALK5 | IPF | Kinase inhibitor |
| + 8 more | Various | ALS, TB, GBM, AD | See models/ |
Full results: models/scientific_discoveries.json
- 99.80% Bell state fidelity (ibm_fez, Feb 2026)
- 5× qubit lifespan extension via QuBIOS Transit Ring
- Real calibration data:
models/ibm_real_results.json
All models were trained on an NVIDIA RTX 3090 Ti (24 GB VRAM) and are available for download from the v1.1.0-models release.
| # | Archive | Size | Source Model | Training Details |
|---|---|---|---|---|
| 07 | macro_news_drug_results.zip | 45 MB | Custom DRL | Crypto/macro news sentiment + drug screening results |
| 08 | medgemma_lora_adapter.zip | 34 MB | MedGemma 4B | LoRA fine-tune r=16, α=32, loss 1.4256 |
| 09 | cxr_foundation_model.zip | 50 MB | CXR Foundation | DenseNet-121, 100% val accuracy |
| 10 | path_foundation_model.zip | 606 MB | Path Foundation | ViT-B/16, 100% val accuracy |
| 11 | brain_mri_model.zip | 167 MB | Custom ResNet-50 | Brain tumor MRI classifier |
| 12 | derm_foundation_model.zip | 15 MB | Derm Foundation | EfficientNet-B0, 99.73% accuracy |
| 13 | medical_training_data_scripts.zip | 1.4 MB | — | Training scripts, datasets, configs |
Total: ~917 MB across 7 archives. See
trained_models/README.mdfor full documentation, usage examples, and loading instructions.
- google/medgemma-4b-it — Medical instruction-tuned foundation model
- google/cxr-foundation — Chest X-ray foundation model
- google/path-foundation — Pathology foundation model
- google/derm-foundation — Dermatology foundation model
QubitPage-OS/
├── README.md ← This file
├── VERSION ← 1.1.0
├── INSTALL.md ← Full installation guide
├── LICENSE ← MIT License
├── requirements.txt ← Python dependencies
├── .gitignore
│
├── src/ ← Core Python backend
│ ├── app.py ← Flask+SocketIO server (main entry point)
│ ├── config.py ← Configuration (all secrets via env vars)
│ ├── qubilogic.py ← QuBIOS engine (Transit Ring, QEC, Teleport)
│ ├── quantum_backends.py ← IBM Quantum + AWS Braket + simulators
│ ├── quantum_kernel.py ← QLang circuit executor
│ ├── quantum_drug_sim.py ← Molecular quantum simulation
│ ├── ai_agent.py ← ARIA AI assistant (Gemini/Groq)
│ ├── gemini_orchestrator.py ← Multi-model AI orchestration
│ ├── med_research.py ← Medical research engine
│ ├── quantummed_routes.py ← API routes for QuantumMed apps
│ ├── report_generator.py ← Discovery report generation
│ ├── user_auth.py ← User auth + per-user API key management
│ ├── test_qubilogic.py ← Core quantum engine tests
│ └── test_qubilogic_extended.py
│
├── templates/ ← Jinja2 HTML templates
│ ├── os.html ← Main OS desktop (all apps, QLang shell)
│ ├── qubilogic.html ← QuBIOS/QubiLogic Memory app
│ └── docs.html ← Documentation wiki
│
├── static/js/ ← Frontend JavaScript
│ ├── quantum-os.js ← Core OS window manager
│ ├── qbp-runtime.js ← QLang browser SDK
│ ├── quantumneuro.js ← QuantumNeuro GBM app
│ ├── quantumtb.js ← QuantumTB research app
│ ├── disease-dashboard.js ← Disease Hub dashboard
│ ├── medlab.js ← MedLab case analysis
│ ├── medfiles.js ← Medical file processor
│ ├── training-viewer.js ← Training results viewer
│ ├── reports.js ← Discovery reports viewer
│ ├── quantumderm.js ← QuantumDerm skin lesion analysis
│ ├── docs-app.js ← In-app docs browser
│ └── real_case_loader.js ← Real patient case loader
│
├── trained_models/ ← Model documentation (archives in GitHub Releases)
│ └── README.md ← Model documentation & usage
│
├── models/ ← Research data & training results
│ ├── README.md ← Data dictionary
│ ├── scientific_discoveries.json ← 13 novel drug candidates
│ ├── ibm_real_results.json ← IBM Fez real hardware results
│ ├── quantum_research.json ← Quantum simulation studies
│ ├── training_metrics_summary.json ← ML training overview
│ └── training_results/
│ ├── comprehensive_drug_screening.json
│ ├── txgemma_admet_full.json
│ ├── fix_results.json
│ └── fix2_results.json
│
├── docs/ ← Extended documentation
│ ├── getting-started.md ← Beginner's guide
│ ├── architecture.md ← System architecture deep-dive
│ ├── medgemma-integration.md ← MedGemma + medical AI guide
│ ├── ai-models.md ← All supported AI models
│ ├── quantum-simulators.md ← Google Cirq, Stim, IBM + more
│ └── api-reference.md ← REST API endpoints
│
└── examples/
├── keys.env.example ← Environment variable template
├── quickstart.py ← Python API quickstart
└── qubitpage-os.service ← systemd service template
| Project | Link | Description |
|---|---|---|
| QLang | github.com/qubitpage/QLang | Quantum Programming Language with 27 native commands, EBNF grammar, and browser SDK |
| QuBIOS | github.com/qubitpage/QuBIOS | Transit Ring quantum middleware — 5× lifespan, 99.80% fidelity |
"The next breakthrough in medicine will not come from a hospital. It will come from a quantum computer."
We are building the first quantum operating system designed specifically to find cures for diseases that have resisted all conventional approaches. Here's why this matters:
- Glioblastoma (GBM): Median survival = 15 months. Zero cures in 50 years.
- ALS (Lou Gehrig's): 100% fatal. 5-year survival = 10%.
- IPF (Lung Fibrosis): Median survival = 3–5 years. No reversal possible.
- Alzheimer's: 55 million patients. All Phase 3 trials have failed.
- Drug discovery: Takes 12–15 years and $2–3 BILLION per drug — most fail.
QubitPage® OS runs on a single computer with a GPU (RTX 3090 or better) and:
| Feature | What It Means |
|---|---|
| Quantum VQE simulation | Simulates molecular binding sites that classical computers can't model accurately |
| MedGemma AI diagnosis | Provides medical-grade diagnostic assistance even in remote areas with no specialist |
| ADMET prediction | Predicts drug toxicity, absorption, and metabolism before any lab test — saving years |
| IBM Quantum integration | Validates results on real 156-qubit quantum hardware (IBM Fez) |
| 13 novel drug candidates | Already discovered via quantum screening (QBP-001 through QBP-013) |
| Remote deployment | One GPU server = full quantum drug discovery lab. No $100M facility needed. |
Imagine a rural clinic in Romania, Ghana, or rural India with:
- One computer (RTX 3090)
- Internet connection
- A doctor
Running QubitPage® OS, that doctor can:
- Diagnose using MedGemma AI trained on millions of cases
- Search quantum drug databases for personalized treatments
- Simulate molecular interactions for the patient's specific genetic profile
- Contribute anonymized research data that improves the global model
This is the future we are building. And we need your help.
We are looking for medical scientists, oncologists, immunologists, virologists, and computational biologists who want to contribute to the most ambitious open quantum medicine project in history.
- Free access to the full QubitPage® OS platform (no credit card needed)
- Access to IBM Quantum real hardware for your research
- Your discoveries published under your name in the
models/scientific_discoveries.json - Collaboration with the Qubitpage® research team
- First access to new drugs and treatments your own research helps discover
📧 Email: contact@qubitpage.com
🌐 Platform: qubitpage.com
🐙 GitHub Org: github.com/qubitpage
Subject line: [RESEARCHER] Your name / Specialty / Institution
We especially welcome researchers from areas with limited access to expensive lab facilities. If you have the knowledge and a computer, QubitPage® OS gives you the quantum lab.
- Platform: qubitpage.com
- Contact: contact@qubitpage.com
- Issues: GitHub Issues
- Org: github.com/qubitpage
QubitPage® — Quantum computing for medicine, starting now.
Copyright © 2026 Qubitpage LIMITED. All rights reserved.
QubitPage® is a registered trademark of Qubitpage LIMITED.
QubitPage® OS is 100% free and open-source. If you find this project valuable, you can help us accelerate quantum drug discovery research.
Our R&D requires significant GPU computing power. We gratefully accept:
| Type | Details |
|---|---|
| 🖥️ GPU Credits | Cloud GPU credits (AWS, GCP, Lambda Labs, RunPod, CoreWeave, etc.) |
| 💰 Donations | Any amount helps fund quantum computing research and development |
| 🔧 Physical GPUs | We accept physical GPU donations (RTX 3090, RTX 4090, A100, H100, etc.) |
Every contribution — whether it's cloud credits, a monetary donation, or even a physical GPU — directly powers the quantum simulations that search for new drug candidates against diseases like Glioblastoma, ALS, Alzheimer's, and IPF.
📧 Contact: contact@qubitpage.com
🌐 Website: qubitpage.com/products/qubitpage-os