This page lists the minimum and recommended system requirements for running RTILA X. Meeting the minimum requirements ensures RTILA X runs. Meeting the recommended requirements ensures a smooth, fast experience.
| Component |
Minimum Requirement |
| Operating System |
Windows 10 (version 1809+), macOS 11 (Big Sur)+, Ubuntu 22.04+ |
| RAM |
4 GB |
| Disk Space |
2 GB free |
| Browser |
Microsoft Edge, Brave Browser, or Google Chrome installed |
| Internet |
Required for initial setup and web scraping |
| CPU |
Dual-core 2.0 GHz |
| Component |
Recommended |
| Operating System |
Windows 11, macOS 14+, Ubuntu 24.04 |
| RAM |
8 GB (16 GB for local AI) |
| Disk Space |
10 GB free (20 GB+ for AI models) |
| Browser |
Microsoft Edge (latest version) |
| Internet |
Broadband connection (10 Mbps+) |
| CPU |
Quad-core 3.0 GHz |
| GPU |
Optional, recommended for local AI (4 GB+ VRAM) |
If you plan to use the Local LLM feature, here are the requirements by model size:
| Model Size |
Minimum RAM |
Recommended RAM |
GPU VRAM |
Notes |
| 7B parameters (Q4) |
6 GB |
8 GB |
Optional |
Good for most tasks. Recommended starting point. |
| 8B parameters (Q4) |
6 GB |
8 GB |
Optional |
Similar to 7B. Slightly better quality. |
| 13B parameters (Q4) |
10 GB |
16 GB |
4 GB+ recommended |
Better quality. Needs more resources. |
| 14B parameters (Q4) |
10 GB |
16 GB |
4 GB+ recommended |
Similar to 13B. |
| 30B+ parameters (Q4) |
20 GB |
32 GB+ |
8 GB+ required |
Best quality. Requires significant resources. |
For a 7B Q4_K_M model at 32K context:
| Component |
Memory |
| Model weights (Q4_K_M) |
~4.5 GB |
| KV cache (F16) |
~2 GB |
| KV cache (Q8_0) |
~1 GB |
| Total VRAM needed |
~5.5–6.5 GB |
Tip: If you don’t have enough VRAM, set GPU Layers to 0 to run on CPU only. It will be slower but will work on any machine.
| Requirement |
Details |
| Version |
Windows 10 (version 1809+) or Windows 11 |
| Server Editions |
Windows Server 2019+ supported. Requires Visual C++ Redistributable (auto-installed). |
| Antivirus |
Some antivirus software may block Playwright. Add an exclusion for RTILA X. |
| Windows Defender |
May show a security warning on first launch. Click “More info” → “Run anyway”. |
| Browser |
Microsoft Edge is pre-installed on Windows 10/11 and is the recommended browser. |
| Firewall |
Ensure Windows Firewall allows RTILA X to access the internet. |
| Sleep Mode |
Disable sleep mode if you have scheduled tasks. Scheduled tasks won’t run if the computer is asleep. |
| Requirement |
Details |
| Version |
macOS 11 (Big Sur) or later. Apple Silicon (M1/M2/M3/M4) and Intel Mac supported natively. |
| Gatekeeper |
On first launch, macOS may block the app. Go to System Preferences → Security & Privacy → “Open Anyway”. |
| Browser |
Install Microsoft Edge, Brave, or Google Chrome. Safari is not supported. |
| Permissions |
macOS may ask for file system permissions. Grant them when prompted. |
| Sleep Mode |
Disable sleep mode if you have scheduled tasks. Go to System Preferences → Energy Saver. |
| Rosetta |
Not required. RTILA X runs natively on both Apple Silicon and Intel. |
| Requirement |
Details |
| Distribution |
Ubuntu 22.04+ recommended. Other distributions with a full Desktop GUI may work. |
| Desktop Environment |
A full Desktop GUI is required. GNOME is recommended. Headless servers are not supported. |
| Dependencies |
Install: sudo apt install libwebkit2gtk-4.0-37 libgtk-3-0 libappindicator3-1 |
| Browser |
Install Microsoft Edge, Brave, or Google Chrome. |
| Permissions |
Ensure the user has read/write permissions to ~/.rtila-x/. |
| Display |
A display server (X11 or Wayland) is required. RTILA X cannot run on headless servers. |
| Requirement |
Details |
| Initial Setup |
Internet connection required for first launch (database initialization, license activation). |
| Web Scraping |
Internet connection required to access target websites. |
| AI (Cloud) |
Internet connection required for OpenRouter and cloud AI providers. |
| AI (Local) |
Internet connection required for initial model download only. No internet needed after download. |
| Public Tunnel |
Internet connection required for the public tunnel (remote execution). |
| CAPTCHA Solving |
Internet connection required for 2Captcha service. |
| Updates |
Internet connection required to check for and install updates. |
| Activity |
Bandwidth |
| Web scraping |
Depends on target website. Typically 1-10 Mbps. |
| AI model download |
4-10 GB initial download. One-time. |
| Cloud AI requests |
Minimal. Text-based requests. |
| Public tunnel |
Minimal. Text-based requests. |
| CAPTCHA solving |
Minimal. Image upload/download. |
| Component |
Size |
| RTILA X application |
~500 MB |
| PocketBase database |
Varies (typically 100-500 MB) |
| Browser profiles |
~100-500 MB per profile |
| AI models (7B Q4) |
~4.5 GB |
| AI models (13B Q4) |
~8 GB |
| Execution logs |
Varies (typically 10-100 MB) |
| Scraped data |
Varies |
| Total (typical) |
2-10 GB |
| Total (with AI models) |
10-20 GB |
| Tip |
Why It Matters |
| Use an SSD |
SSDs significantly improve database and model loading performance. |
| Close unnecessary applications |
Free up RAM for RTILA X and the browser. |
| Use headless mode for scheduled tasks |
Headless mode uses less memory and is faster. |
| Use a smaller context size |
Smaller context sizes use less memory and are faster. |
| Use Q8_0 KV cache |
Reduces memory usage with negligible quality loss. |
| Use a wired internet connection |
More stable than Wi-Fi for web scraping. |