Managed Local AI
Your own private AI running on your GPU, with Open WebUI — a ChatGPT-style chat window for your team.
- Ollama standard
- Open WebUI default
- No third-party AI API required by default
- Managed setup and updates
Managed Local AI on your own GPU
Your own private AI — like ChatGPT, but it runs on your server and your data never leaves the house. We build it on Ollama (the open-source app that runs AI models on your own machine), keep it updated, and show you real test results before you pay for performance.
- Private AI built on Ollama, the open-source standard
- Open-source hosting managed with CyberPanel
- Domains, support, and clear ways to get started
Runs on Ollama with Open WebUI, a private ChatGPT-style chat for your team. Outside AI services are used only if you ask for them.
AI Apps
Private document search, team chat, workflow automation, and creative image tools — all open-source and managed by us.
- AnythingLLM or LibreChat
- Flowise and n8n options
- ComfyUI for creative GPU workflows
- vLLM only as optional advanced layer
GPU Infrastructure
The right GPU server for running AI, generating images, and private automation — sized to what you actually need.
- Dedicated setup
- Storage and backup planning
- Monitoring and maintenance
- Benchmark before performance promises
Open Source Hosting
Managed open-source hosting with CyberPanel, domains, SSL, DNS, and human support.
- CyberPanel control panel
- WordPress, Nextcloud, Matomo and more
- No cPanel or CentOS claims
- Built for long-term maintenance
Domains
Domain registration, renewal, transfer guidance, and DNS support for open-source projects and teams.
- Popular TLDs with USD pricing
- Renewal notes shown clearly
- Transfers reviewed by registry rules
- DNS basics included
Your private AI, set up and kept running
We install it, secure it, keep it updated, and explain how it works. By default your AI runs entirely on your own server, with no outside AI service needed.
Open WebUI
The chat window your team uses in the browser — like ChatGPT, but private to your server.
AnythingLLM
Lets your AI answer from your own documents, when that fits what you need.
vLLM optional
An advanced speed option we only add after testing it on your model and GPU.
Connect your own tools to your private AI
Many teams don't need a new chatbot first. They need their existing software to talk to a private AI, with access kept under control. We test that connection before anyone relies on it.
Works like the OpenAI API, but local
Your private AI can speak the same "language" as the OpenAI API, so your existing scripts and apps can point to it instead — once we've checked it's ready.
Who can see and do what
For teams, we set up user roles (RBAC), single sign-on (SSO/OIDC), API keys, and per-model permissions before anyone goes live.
Extra speed only after fit checks
There are advanced ways to make models smaller and faster on the GPU (AWQ, GPTQ, GGUF, FP8). We treat them as an optional test track, never an upfront promise.
Our promise: we test the connection on your own server and show you the results before you rely on it in production.
Private AI agents, tested safely first
An AI "agent" can take actions for you, not just chat. That is powerful, but it adds risk. Our Business Secure plan tests each agent on a fixed set of real tasks first, instead of promising a hands-off robot assistant.
gpt-oss 20B as a benchmark candidate
OpenAI positions gpt-oss models for reasoning and tool-use workflows, while Ollama lists the 20B path around 14 GB with a 128K context window (how much text it can read at once). On this server class, that remains a measured Business Secure trial.
OpenAI gpt-oss-20b model card →Ollama gpt-oss →gpt-oss-safeguard 20B policy benchmark
OpenAI describes gpt-oss-safeguard as open safety reasoning models for custom safety policies. Treat the 20B model as a measured policy-classification benchmark for local agent workflows, not an automatic moderation guarantee.
OpenAI safeguard announcement →Hugging Face safeguard-20b →Tools are treated as privileged access
Open WebUI documents that tool access can execute arbitrary Python code. We scope tool permissions, API keys, and provider credentials before any agent workflow touches customer data or business systems.
Open WebUI permissions →Fixed-task audit before rollout
The benchmark uses a fixed task set, allowed tools, blocked tools, sample data, logging expectations, and rollback notes. Production use waits for GPU/Ollama health and a repeatable target-workflow smoke test.
Our promise: we prove an agent works safely on a fixed set of real tasks, on your own server, before we ever call it production-ready.
First we check it works, then we pick your model
Local AI changes fast. We turn that uncertainty into a short, practical set of checks before you commit to a bigger rollout.
Runtime check
We verify GPU drivers, CUDA visibility, Ollama or container service health, storage, backups, and secure access before model work starts.
20 GB model shortlist
For RTX 4000 Ada class systems, we shortlist realistic small-to-medium models such as Qwen, Gemma 4, or DeepSeek distill variants and document tradeoffs.
Benchmark report
Performance claims are made after testing the target model, quantization, context length, and user workflow. If vLLM is useful, it stays an optional advanced layer.
Which AI models fit a 20 GB card right now
Checked against current sources on 2026-06-02. These are models we would test on a 20 GB card — candidates, not guaranteed speeds.
Qwen3 Embedding + Reranker
Qwen3-Embedding and Qwen3-Reranker 0.6B/4B/8B cover multilingual retrieval, code search, and source ranking. Test corpus quality, storage, latency, and answer citations before rollout.
Qwen3 Embedding announcement →Qwen3-Embedding-4B card →Qwen3-VL + Docling baseline
Qwen3-VL 8B is a current OCR and document-structure candidate; Docling/OCR gives a deterministic parsing baseline for PDFs, tables, reading order, and field-level checks.
Qwen3-VL-8B model card →Qwen3-VL Embedding + Reranker trial
Qwen3-VL-Embedding-8B and Qwen3-VL-Reranker-8B are current multimodal retrieval candidates for text, images, screenshots, videos, and mixed documents. On 20 GB systems, scope them as a measured visual-RAG benchmark with corpus size, vector storage, latency, and fallback limits.
Qwen3-VL-Embedding-8B card →Qwen3-VL-Reranker-8B card →Qwen3-Coder 30B benchmark
Ollama lists the 30B path around 19 GB with 256K context, so on a 20 GB GPU it is benchmark-only with strict context, concurrency, and fallback limits.
Ollama Qwen3-Coder listing →Gemma 4 E4B/26B trial
Gemma 4 E4B is the low-memory assistant and multimodal candidate. Gemma 4 26B and 31B are benchmark-only on 20 GB because Ollama lists 18 GB and 20 GB footprints; concurrency, context, and MTP latency gains stay gated by local smoke tests.
Google Gemma 4 announcement →Gemma 4 MTP drafters →gpt-oss 20B private reasoning benchmark
OpenAI positions gpt-oss-20b for local and specialized reasoning use, and Ollama lists the 20B path around 14 GB with 128K context. Treat it as a strong Business Secure benchmark candidate with strict context, tool-use, safety, and latency checks before rollout.
OpenAI gpt-oss-20b card →OpenAI model card PDF →Our promise: we make no speed or performance claim until the drivers, Ollama, and your chosen model all pass tests on your real server.
Four ready-made private-AI packages
Four practical packages for a 20 GB GPU. Each one starts with a real test on your server (a "smoke test") before we make any performance claim.
Private RAG with sources
An AI that answers using your own documents and shows its sources — this is called RAG. Good for internal files, support archives, policies, and project knowledge, where every answer should point back to where it came from.
- Open WebUI or lightweight team UI
- Qwen3 or Gemma 4 chat candidate
- EmbeddingGemma or Qwen embedding trial
Smoke test: ingest a representative corpus, ask fixed benchmark questions, require cited answers, and record VRAM, latency, and miss behavior.
Start Team RAG benchmark →Local PDF and image extraction
For invoices, forms, screenshots, and operational documents that need local extraction support without sending files to a cloud AI API by default.
- Qwen3-VL/Qwen2.5-VL benchmark
- Docling/OCR fallback for hard scans
- Field-level error notes
Smoke test: run real sample pages against expected fields, measure false positives, unsupported layouts, throughput, and VRAM peak.
Read document intake benchmark →Request document workflow trial →Visual RAG and evidence search
For teams that need to retrieve answers from screenshots, diagrams, scanned pages, product images, or short video captures with evidence links instead of plain text-only RAG.
- Qwen3-VL-Embedding 8B recall trial
- Qwen3-VL-Reranker 8B ranking trial
- Source thumbnails and miss analysis
Smoke test: index a small mixed-media corpus, ask fixed visual-search questions, record top-k misses, reranker gains, storage size, VRAM, and latency before rollout.
Start visual RAG benchmark →Local transcription and meeting notes
For interviews, internal meetings, and support recordings where private audio handling and predictable operations matter more than a generic SaaS workflow.
- Whisper or faster-whisper benchmark
- English and German sample set
- Optional local summary pass
Smoke test: transcribe representative 5 to 30 minute files, record runtime factor, language quality, segmentation limits, and GPU use.
Scope transcription benchmark →Our promise: these packages are sold as setup and ongoing management. Any live performance claim waits until it is tested on your actual server.
RAG Retrieval Quality Audit
A smaller, paid first step. For teams that want an AI to search their own documents but aren't yet sure their files, their questions, and their access rules are ready for a full monthly rollout.
Representative corpus first
We start with a limited set of real documents, screenshots, policies, tickets, or manuals and turn them into a fixed retrieval benchmark instead of ingesting everything blindly.
Qwen3 small-to-medium retrieval path
Qwen3-Embedding 0.6B/4B and Qwen3-Reranker 0.6B/4B are the default audit candidates. The 8B path stays optional when corpus size, latency, and the 20 GB VRAM budget justify it.
Answer quality before rollout
The result is a short report with top-k misses, citation quality, reranker gains, storage notes, privacy boundaries, and a clear decision: fix sources, run a pilot, or move to Team RAG.
Our promise: this audit gives you real evidence of search quality and a clear next step. It makes no live-performance claim until the GPU, Ollama, and your chosen models pass tests on your own server.
Plans built around real work
Use this page to pick the right starting point, then confirm the exact setup, renewal terms, and any test notes at checkout.
BYO Server Management
from $299.18/mo
Already have your own GPU server or rented one? We manage Ollama, Open WebUI, updates, and monitoring on it. (BYO means bring your own server.)
- Driver and Ollama readiness check
- Open WebUI installation
- Hosting costs paid directly by you
Local AI Managed
from $699.42/mo
For private local model hosting with a managed operating layer and no third-party AI API required by default.
- 20 GB model-fit shortlist
- Benchmark report before promises
- Managed updates, monitoring, backups
Team RAG
from $999.60/mo
For teams that need document-assisted local AI, curated embeddings, user roles, and clearer knowledge workflows.
- Knowledge/RAG setup
- Embedding and model guidance
- Prioritized support
Business Secure
from $1,499.90/mo
For production business use where security, audit preparation, update windows, and support scope matter.
- Security hardening
- OIDC/SSO preparation
- Monthly review and change windows
Before we make any live-AI claim, we check the GPU drivers and confirm Ollama is healthy. A 20 GB RTX 4000 Ada card is great for small-to-medium models, not the very largest frontier models.