MoUtils | The Local AI & Utility Toolkit
MoUtils 1.0 is now available

The ultimate toolkit.
Purely local.

A comprehensive utility toolkit featuring native desktop tools, offline file processing, and a powerful on-device multimodal AI agent. Total privacy. Zero cloud dependency.

Try Free in Browser →Download DesktopExplore Capabilities
Intelligence

Multimodal AI.
Zero latency.

Experience the power of language, vision, and audio models running natively on your hardware. No API keys. No subscriptions. Absolute privacy.

15+ Models

Text Intelligence

Engage with leading open-weight LLMs like Llama 3, Qwen, and Gemma. Perfect for coding, writing, and complex reasoning.

FastVLM & Gemma

Vision Analysis

Analyze images, extract text, and understand visual context locally using state-of-the-art vision-language models.

Real-time

Audio Processing

Transcribe meetings, dictate notes, and process speech instantly with entirely on-device audio models.

Architecture

Designed for sovereignty.

By bypassing traditional cloud infrastructure, MoUtils delivers uncompromising data sovereignty with general-purpose offline tools and a powerful local agent layer.

Offline-First Tooling

Native developer, file, and system utilities that work entirely offline without cloud dependencies or telemetry.

Agentic Architecture

Local AI models can execute system tools directly, acting as an autonomous assistant for your computer.

Native Performance

Built with platform-specific APIs to ensure minimal resource footprint and maximum hardware utilization.

Privacy by Design

All file processing, network intelligence, and AI inferences happen locally. Your data never leaves your device.

Instant Access

Free Web Toolkit.
Zero installation.

Try core MoUtils utilities directly in your browser. Fully client-side, 100% private, and completely free to use.

Preview Web App →

Text Converter & Hasher

Encode, decode, and generate cryptographic hashes (Base64, URL, HTML, MD5, SHA-256) entirely client-side.

Side-by-Side Diff Compare

Compare files and code snippets with live syntax highlighting and fine-grained character-level differences.

Network Utilities & Diagnostics

Inspect ISP speeds, query hardware MAC manufacturers, and audit public IP data safely in your browser.

Library

Scaled for your hardware.

A curated selection of models spanning text, vision, audio, and function calling. Seamlessly downloaded and managed on-demand.

Gemini Nano (System OS)
VisionTools
512 MB RAM • 0 MB
Apple Foundation Models (System OS)
VisionTools
512 MB RAM • 0 MB
SmolLM 135M Instruct
Text
768 MB RAM • 142 MB
LFM2.5 230M
Thinking
768 MB RAM • 168 MB
FunctionGemma 270M Mobile Actions
Tools
896 MB RAM • 275 MB
Gemma 3 270M IT
Text
896 MB RAM • 307 MB
Qwen 2.5 0.5B Instruct
Tools
1 GB RAM • 512 MB
FastVLM 0.5B
Vision
1 GB RAM • 512 MB
Qwen 3 0.6B
ToolsThinking
1.1 GB RAM • 586 MB
Gemma 3 1B IT
Tools
1 GB RAM • 520 MB
Qwen 2.5 1.5B Instruct
Tools
2.1 GB RAM • 1.60 GB
DeepSeek R1 Distill Qwen 1.5B
ToolsThinking
2.3 GB RAM • 1.70 GB
Gemma 3n E2B IT
VisionToolsAudio
3.6 GB RAM • 3.03 GB
Gemma 4 E2B IT
VisionToolsThinkingAudio
3 GB RAM • 2.40 GB
Qwen 3 4B
ToolsThinking
3 GB RAM • 2.48 GB
Phi-4 Mini Instruct
Tools
4.5 GB RAM • 3.90 GB
Gemma 4 E4B IT
VisionToolsThinkingAudio
4.9 GB RAM • 4.30 GB
Qwen 3 8B
ToolsThinking
5.1 GB RAM • 4.55 GB
Gemma 4 12B IT
ToolsAudio
6.6 GB RAM • 6.10 GB
Utilities

Focused capabilities.

Built-in operations for every workflow, eliminating context switching and the need for scattered web tools.

Windows Tools

  • Windows Cleaner & OptimizerDeep cleanup and system optimization for Windows
  • Software ManagerDiscover, install, update, and remove software easily using Winget

Developer Tools

  • Text Diff CompareSide-by-side text comparison and merging
  • Database ViewerInspect offline (SQLite) and online (Postgres/MySQL) databases
  • Converter & HasherEncode or decode text and calculate or verify text/file checksums
  • JSON InspectorValidate, format, and minify JSON locally

Network Tools

  • Network ScannerScan devices on the current local network subnet
  • Port ScannerScan open ports on any target IP or domain
  • Ping ToolSend ICMP echo requests to a target IP or domain
  • DNS & Reverse LookupQuery A/AAAA records for a domain, or PTR records for an IP
  • ISP & Speed TestCheck public IP/ISP details and test network speeds
  • IKWYD CheckerLook up torrent downloading history associated with an IP

AI Tools

  • Local ChatMulti-turn local chat using your active on-device model
  • Local AgentTask-oriented on-device agent with function calling

AI Text Prompts

  • Local Text TranslatorTranslate text between selected languages with your local model
  • Grammar PolisherImprove grammar, flow, and style
  • Email RewriterRewrite emails in different tones
  • Text SummarizerSummarize long text into concise summaries
  • Code ExplainerExplain code in plain language

AI Vision Prompts

  • Receipt DigitizerOCR + AI to extract receipt fields
  • Calorie EstimatorEstimate ingredients and calories from photos
  • Scene DescriberGet a detailed description of any image
  • Document ScannerExtract text from documents, whiteboards, or handwritten notes

AI Audio Prompts

  • Audio TranscriberTranscribe speech from an audio file to text
  • Audio SummarizerTranscribe and summarize the content of an audio recording
  • Meeting Action ItemsExtract tasks and action items from a meeting or voice memo
FAQ

Common Questions

You only need an internet connection to download the app and the AI models initially. Once downloaded, all inferences and utility tools run 100% locally on your machine.
No. MoUtils does not collect telemetry or usage analytics. Crash reports (excluding prompts, files, or sensitive input) are reported via Sentry to diagnose bugs and improve app stability. Your prompts, files, and usage data never leave your device.
MoUtils itself is lightweight, but AI models require RAM. We recommend at least 8GB of RAM for smaller models (like Qwen 0.5B), and 16GB+ for larger models (like Gemma 4 12B). A dedicated GPU will significantly improve inference speed, but CPU inference is fully supported.
The AI agent is restricted to a specific set of safely exposed tools (like managing windows, viewing system stats, or editing specific files). It cannot execute arbitrary code unless explicitly given a sandboxed terminal tool by the user.
Pricing

Simple, transparent pricing.

Pay once, own forever. No recurring token costs or cloud subscriptions.

Lifetime License

A single-time purchase for a lifetime of local AI utilities.

$5