Websites that load faster than you can click
Under 0.3 seconds load time. Local AI integration. Pay once, maintain yourself.
Why Simple Network?
Load time under 0.3 seconds
While other websites take 2–5 seconds to load, yours is already done.
Load time under 0.3 seconds
While other websites take 2–5 seconds to load, yours is already done.
No WordPress, no website builders — statically generated with Astro. That means: no server querying databases on every request. Your site is built at compile time and served as ready-made HTML. The result: load times that Google rewards and visitors appreciate.
Local AI instead of cloud dependency
Your AI runs on dedicated Hetzner hardware. No cloud APIs, no third-party dependency.
Local AI instead of cloud dependency
Your AI runs on dedicated Hetzner hardware. No cloud APIs, no third-party dependency.
We use llama.cpp for CPU-optimised inference with GPT-OSS-20B and Qdrant as a vector database for RAG. The backend runs in a separate network — no access to your company network. You pay once for setup and development, then only the server costs directly to the host.
One-time instead of subscription
WordPress agencies charge €20–50 per month. With us, you pay once.
One-time instead of subscription
WordPress agencies charge €20–50 per month. With us, you pay once.
After handover, the website is yours. The code is documented, training is included. Pure website hosting costs only the domain (~€20/year) and a small server. AI packages have transparent server costs directly at Hetzner — no licence fees, no plugin subscriptions, no hidden costs.
Ready for AI browsers
Structured data and Schema.org make your site visible to the next generation of search engines.
Ready for AI browsers
Structured data and Schema.org make your site visible to the next generation of search engines.
AI-based browsers like ChatGPT Search or Perplexity don't just read text — they need structured data. Every page has Schema.org JSON-LD, semantic HTML, and high information density. So you're not only found on Google, but also referenced by AI.
Packages
Pay once. Maintain yourself. Server costs paid directly to the host.
Website
€3,500 one-time High-performance Astro website with under 0.3s load time — no AI
Website
€3,500 one-timeHigh-performance Astro website with under 0.3s load time — no AI
- Statically generated with Astro (SSG)
- Load time under 0.3 seconds
- WCAG 2.2 AA accessible
- Schema.org & SEO optimised
- GDPR compliant, fully cookie-free
- Bilingual DE/EN with hreflang
- Documented code, training & support
Basis AX42
€5,000 one-time Website + AI webchat on dedicated Hetzner server (AMD Ryzen)
Basis AX42
€5,000 one-timeWebsite + AI webchat on dedicated Hetzner server (AMD Ryzen)
- Everything from Website package
- AMD Ryzen 7 PRO 8700GE, 64 GB DDR5
- RAG webchat with GPT-OSS-20B (llama.cpp)
- BGE-M3 embedding + Qdrant vector database
- Up to 3 parallel chat sessions
- 16K token context window
- Separate network — no access to company network
- ~€60/month server costs (Hetzner, direct)
Pro AX162
€7,000 one-time Webchat + backend AI with LoRA adapters on AMD EPYC server
Pro AX162
€7,000 one-timeWebchat + backend AI with LoRA adapters on AMD EPYC server
- Everything from Basis package
- AMD EPYC 9454P (48 cores), 256 GB DDR5 ECC
- Separate backend AI with LoRA fine-tuning
- User accounts per employee
- Autonomous server maintenance via AI assistant
- Up to 6 parallel sessions, 64K context
- Security architecture: read-only + secured folder
- ~€250/month server costs (Hetzner, direct)
Enterprise GEX131
€10,000 one-time GPU-accelerated AI with telephone integration on NVIDIA RTX PRO 6000
Enterprise GEX131
€10,000 one-timeGPU-accelerated AI with telephone integration on NVIDIA RTX PRO 6000
- Everything from Pro package
- NVIDIA RTX PRO 6000 Blackwell, 96 GB GDDR7 ECC
- Intel Xeon Gold 5412U (24 cores), 256 GB DDR5 ECC
- GPU-accelerated inference — minimal response time
- Telephone AI for customer interaction (8B model, real-time)
- Maximum reasoning depth for frontend & backend AI
- ~€900/month server costs (Hetzner, direct)
Frequently Asked Questions
Why is the site so fast?
The website is statically generated at build time — no server needs to query databases or execute PHP on every request. The finished HTML files sit directly on the server and are delivered instantly.
What does the website cost after handover?
For the website only: just the domain (~€20/year) and hosting (~€5/month). With AI packages, the website runs directly on the booked server — no separate hosting needed. You pay the server costs directly to the host (Basis ~€60, Pro ~€250, Enterprise ~€900/month at Hetzner). No licence fees, no plugin subscriptions, no hidden costs. The code belongs to you.
Can I edit the site myself?
Yes. The code is documented and an introduction is included. For text changes, basic Markdown skills are enough. From the Pro package, the backend AI assists you directly — it knows the code and server configuration. For larger extensions or changes, I'm available any time as a commissioned project.
Why local AI instead of ChatGPT API?
No recurring API costs, full control over your data, and no dependency on third parties. The AI runs on dedicated Hetzner servers (AMD Ryzen/EPYC or NVIDIA RTX PRO 6000) in a separate network — no access to your company network. From the Pro package, the AI gets LoRA fine-tuning and can even assist with server maintenance.
What does WCAG 2.2 AA mean?
This is the European accessibility standard. From 2025, accessible websites are legally required for many businesses (European Accessibility Act). Our sites meet this standard from the start.
Does the site work with AI browsers?
Yes. Every page has structured data (Schema.org), semantic HTML, and high information density. This makes it optimally readable for AI-based search engines like ChatGPT Search or Perplexity.