OwnAI
own the AI and keep your data private
Run local AIs on your hardware and experience the OwnAI agent with 140+ native tools and skills. Create beautiful PPTX, Word, Excel, and HTML websites and dashboards — or generate images and videos via optional OpenRouter cloud models. Anonymization functions are in place in case you use cloud AIs via APIs.
No matter what - your data stays yours.
- 30-Day Free Trial
- One-Time Payment
- 1 Device per License
- No Subscription
OwnAI in Action
OwnAI look and feel
Take a closer look at the OwnAI interface and discover its convenient features—export to Word or PDF, edit responses, save to long-term memory, and explore even more capabilities.
OwnAI video demos
Watch OwnAI in action and experience the power of local, private and intelligent AI. Watch our tutorials to get started quickly with OwnAI — or dive deeper with power-user guides.
Own 140+ agent tools
OwnAI offers 140+ built-in tools in Agent mode. Manage your projects, folders, and personal files; create modern, visually appealing dashboards, diagrams, presentations and reports in HTML, Powerpoint, Word, Excel. ODT etc. - your hardware processes your data.
Over 50 reasons for OwnAI
Every feature is designed around one principle: you own your data, technology and privacy.
Privacy & Security
- Local-first by default — with local AI models your data never leaves your computer
- Optional Cloud AI via OpenRouter — anonymize sensitive data before it leaves your machine
- Encrypted credential storage with PBKDF2-SHA256 (200k iterations)
- No cloud required for core use — local models keep working without internet
- Sandboxed coding agent with Safe / Balanced / Aggressive profiles
- Verification-skipped detector warns you before silent bugs slip through
Coding Agent
- Full local coding agent with 140+ tools
- MCP support — scale OwnAI to fit your needs
- Cursor-compatible agent skills (SKILL.md) — load workflows for coding, docs, and media
- Native GitHub integration — read, edit, and commit under configurable guardrails
- Run tests, linters, formatters — all from within the AI conversation
- Plan-before-execute mode for full visibility before changes are made
- Vision support — upload screenshots or let the agent capture and analyze the screen
- Image and video generation tools via OpenRouter when cloud models are configured
- Code execution capability to verify generated code works
- Fully modifiable agent output modes and composer pane for full agent execution transparency
Discord Bridge
- Talk to your OwnAI agent from Discord — same local or cloud models as the desktop app
- Encrypted Discord bot token in Settings; optional auto-start with OwnAI
- Slash commands for status, plan, image, video, search, memory, and collections
- Approve / Deny agent confirms via Discord embeds when guardrails require it
- RAG, long-term memory, tools, image generation, and video generation available through the bridge
Cost & Independence
- 30-day free trial — then 9.99 € one-time for one license (one device), no recurring fees
- Independence from OpenAI, Google, Anthropic pricing & ToS changes when using local models
- Leverages your existing hardware — NVIDIA and AMD GPUs or integrated CPU (MAC support is in development)
- Open ecosystem: HuggingFace models, Piper TTS, Whisper, ChromaDB
Document Analysis & RAG
- Analyze PDFs, DOCX, XLSX and images with vector-based retrieval
- Multi-stage processing for documents exceeding context limits
- Chunk-based vector retrieval with ChromaDB similarity search
- Up to 10 attachments per query with automatic token budget management
- Data-safe document analysis — with local models confidential data never leaves your drive
Audio & Video — TTS & STT
- Piper text-to-speech on CPU — natural voices, fully offline
- Whisper speech-to-text with automatic language detection
- Transcribe local audio to text (wav, mp3, m4a, and more) — fully offline with Whisper
- Transcribe local video files to text via ffmpeg + Whisper
- YouTube video transcription — captions when available, Whisper fallback, then summarize
- Dedicated Transcription tab for YouTube and video workflows with progress and saved transcripts
- Streaming audio output — hear responses as they're generated
- Audio waveform visualization during speech-to-text; auto-language & voice switching
Memory & Continuity
- Long-term memory stored locally — the AI remembers important details across sessions
- Built-in Memory Manager to review, update, and forget stored facts
- @-pins and attach context — pin files into the conversation when the agent needs them
Exports & Deliverables
- Create Word, Excel, PowerPoint, PDF, HTML, ODT, and RTF from the agent
- Export AI responses, dashboards, diagrams, and reports directly from the app
- Modern HTML websites and interactive dashboards generated on your hardware
AI Model Flexibility
- Choose the best model for every job — security, creative, coding, math
- Dual local formats: Safetensors (GPU) + GGUF (CPU)
- Optional OpenRouter cloud models — switch between local and cloud from the GUI
- Image generation via OpenRouter — from chat mode or Discord `/image`
- Video generation via OpenRouter — text-to-video and image-to-video; Discord `/video`
- Built-in model manager — download & switch local models from the GUI
- Per-model context window configuration for memory optimization
- GPU Memory Guard prevents WDDM paging crashes on Windows
Video Generation
- Text-to-video and image-to-video via optional OpenRouter cloud models
- Composer toggle in Chat, Documents (RAG), and Memory modes — describe a clip and send
- Agent tool
generate_video— ask the coding agent to create short videos - Discord
/video(API mode) — same pipeline as the desktop app - Cloud AI Manager defaults — video model, aspect ratio, duration, optional audio
- Clips saved locally under your media folder and playable in-app
- Optional confirm dialog in Agent Manager before spending provider credits
Web Search
- Integrated DuckDuckGo web search extends knowledge beyond training cutoff
- Source context with citation — see where information came from
- Full control over web search — toggle the tool on or off directly in chat mode
Automation & Local API
- Local OpenAI-compatible API server — connect scripts, tools, and the Discord bridge
- Recurring agent jobs — schedule daily summaries, reviews, or data analysis on a timer
- Jobs run locally; use local models offline or optional OpenRouter for cloud jobs
Customization & Control
- Over 300 top-level configurable settings — fine-tune every aspect of OwnAI
- Custom system prompts per mode (Chat, Agent, Documents etc.)
- Save favorite prompts as one-click templates
- Token status bar and token history dashboard — see exactly how tokens are allocated
OwnAI vs Cloud Alternatives
| Feature | OwnAI | Cloud AI APIs |
|---|---|---|
| Cost | 30-day free trial, then 9.99 € one-time 1 license, 1 device |
monthly subscription costs |
| Privacy | Local-first by default Data stays on your PC with local models |
Your data sent to external servers and is used as training data |
| Latency | Instant (local models) | 200-1500ms round-trip |
| API Key | Not required for local models Optional for OpenRouter |
Mandatory (security risk) |
| Cloud Login | None for local use Optional OpenRouter only |
Required for usage |
| Hidden Costs | Zero for local use OpenRouter billed separately if you enable it |
Sudden price hikes, API limits |
| Data Security | PBKDF2 + XOR Encrypted | Depends on vendor security |
| Image & Video Gen | Optional via OpenRouter Text/image-to-video locally saved Confirm before spend |
Usually subscription-only or pay-per-generation APIs |
System Requirements
Choose a configuration that matches your hardware. Local models run fully offline using safetensors or llama.cpp GGUF quantization. Optional OpenRouter cloud models need no local GPU.
| Category | 🖥️ CPU Only | 🎮 8 GB GPU | 💻 16 GB GPU | 🚀 24 GB GPU | 🔥 32 GB GPU |
|---|---|---|---|---|---|
| Suggested GGUF Model | Qwen3.5-4B (Q5_K_M) |
Qwen3.5-9B (Q4_K_M) |
Qwen3.5-9B (Q4_K_M) |
Qwen3.6-27B (Q3_K_M) |
Qwen3.6-27B (Q5_K_M) |
| Min RAM | 16 GB | 16 GB | 24 GB | 32 GB | 32 GB |
| Min Storage | 20 GB SSD | 20 GB SSD | 25 GB SSD | 30 GB SSD | 30 GB SSD |
| Context Window | ~8K tokens | ~32K tokens | ~90K tokens | ~90K tokens | ~128K tokens |
💡 Download local AI models from HuggingFace through OwnAI's built-in model manager — or connect optional OpenRouter cloud models in Settings.
Download OwnAI 1.5
30-day free trial, no registration required. After the trial phase, purchase OwnAI 1.5 for only 9.99 € — one license for one device, one-time payment, no recurring fees. Own and don't rent the AI technology!
❓ Frequently Asked Questions
Do I need a powerful GPU?
No, but a GPU is strongly recommended for local models. CPU-only inference is significantly slower — especially on Windows and Linux — and may not be practical for larger models. For a smooth local experience we recommend a dedicated GPU with at least 8 GB VRAM and 16 GB system RAM. The app will automatically use your available resources if no GPU is present. Prefer cloud models? Optional OpenRouter works without a local GPU.
This website was crafted entirely with OwnAI — powered by Qwen3.6 27B GGUF in Agent mode on an NVIDIA RTX 4090 (24 GB VRAM) 😊
Which models are supported?
OwnAI supports local LLMs including DeepSeek V4, Qwen 3.6, Llama 3, Mistral, Gemma 4, and many more via Safetensors and GGUF (Llama.cpp / Transformers). Optionally, connect OpenRouter to use cloud models from providers such as OpenAI, Anthropic, Google, and others — without leaving the OwnAI UI. Text-to-text, speech-to-text, text-to-speech, image-to-text, image generation, and video generation (text-to-video / image-to-video) are supported in OwnAI 1.5.0.
Can OwnAI generate videos?
Yes — with an optional OpenRouter API key and a configured cloud video model (Cloud AI Manager → Video model). Toggle the video-generation button in Chat, Documents, or Memory mode and describe the clip; attach an image for image-to-video. In Agent mode, ask the agent to create a video (`generate_video`). From Discord (API mode), use /video …. Clips are saved under your local media folder and can be played in-app. OpenRouter / provider usage is billed separately.
Is this really 100% offline?
With local models: yes. OwnAI runs completely on your machine and no chat data leaves your computer. Optional features that use the network — DuckDuckGo web search and OpenRouter cloud models — are off unless you enable them. You stay in control: disable web search and stick to local models for a fully offline workflow.
What is OpenRouter and do I need it?
OpenRouter is an optional cloud gateway to many AI providers (chat, image generation, and video generation). You do not need it — OwnAI works fully with local models. If you want stronger or specialized cloud models without a powerful GPU — including text-to-video / image-to-video — add your OpenRouter API key in Settings and pick models in Cloud AI Manager. OwnAI can anonymize sensitive data before it is sent. Local use never requires an API key or cloud account.
How long is the free trial and what happens after?
You get 30 days of full access — completely free, no credit card required. After the trial, OwnAI costs 9.99 € one-time for one license covering one device. No recurring fees, no subscription. OpenRouter cloud usage (if you enable it) is billed separately by OpenRouter / the model provider.
Can I switch between models without restarting?
Yes! Model switching works instantly via the Settings Expert view. No restart required — OwnAI loads the new model on the fly.
What is RAG and how does it work in OwnAI?
RAG (Retrieval-Augmented Generation) indexes your local documents so the AI can provide context-aware answers based on your own files. Indexing and retrieval run fully offline on your machine. When you use a local LLM, no document content leaves your PC; with optional OpenRouter, only the context you send in a prompt may leave — and you can anonymize sensitive data first.
How does long-term memory work?
OwnAI stores conversation context across sessions locally on your machine. The built-in sentence transformer embeds key information so the AI remembers important details from previous conversations — no cloud sync required.
Can I use OwnAI with my GitHub repositories?
Yes! OwnAI has native GitHub integration. The agent can read, edit, and commit code directly to your repositories — all controlled by configurable guardrails.
What are guardrails and how do I configure them?
Guardrails are safety switches in the Agent Manager that control which actions the agent can execute autonomously. You can set each tool to "Confirm" (requires your approval) or "Allow" (runs automatically).
What is Plan Mode?
In Plan Mode, the agent first drafts a step-by-step execution plan before taking any action. You review and approve each phase — giving you full transparency and control over what the agent does.
What is MCP and how do I use it?
MCP (Model Context Protocol) lets you connect external tools and services to extend OwnAI's capabilities. Configure MCP servers in the Settings Expert view to scale OwnAI to fit your needs.
Can I schedule recurring agent jobs?
Yes, the built-in job scheduler lets you run tasks on a timer — for example daily summaries, automated code reviews, or periodic data analysis. Scheduling runs locally; use local models for a fully offline workflow, or optional OpenRouter if you prefer cloud models for those jobs.
What export formats are supported?
OwnAI supports PDF, Word (.docx), Excel (.xlsx), PowerPoint (.pptx), RTF, ODT, HTML, and more. Export AI responses, dashboards, and reports directly from the app.
How many settings can I customize?
Over 300 top-level settings in the Settings Expert view — from model parameters and token limits to UI themes, guardrails, and MCP configuration. Tailor OwnAI to your exact needs.
What's the difference between GGUF and SafeTensor?
GGUF is optimized for CPU/GPU inference via llama.cpp and is the most common format for local models. SafeTensor (Safetensors) is HuggingFace's secure model format. OwnAI supports both.
Can multiple users share one license?
One license covers one device (9.99 € one-time). Need more devices? Purchase an additional license — no recurring fees.
💬 Feedback for OwnAI 2.0
Help us shape the next version! Share your feature requests, improvements, and ideas.