⬢ v2 'Tezcat' — Major Release · Linux + Windows

Bundled model: Qwen 3.5 ships with the installer — open it and it works. No model download dance.
CUDA acceleration: 3–15× faster on NVIDIA GPUs. Auto-detects. Falls back to CPU if no GPU.
Developer toolbox: SHA hash checker, CSS/JS minifier, JSON validator/formatter — all inside the app.
Calendar agent: "Schedule Pedro Tuesday 3pm" — books the visual calendar, WhatsApps Pedro automatically.
Voice on GPU: Whisper.cpp with CUDA — 0.3s transcription on RTX 3060.
Multi-channel daemons: WhatsApp, Telegram, Email, Slack, Discord — all from one Service Switch panel.

✦ v2.1 in development — image generation, document editing & CPU-light mode. See what's coming →

⬢ TEZCAT · v2.0 · LIVE ON LINUX + WINDOWS

It just works.
The moment you open it.

v2 ships with the model bundled. No setup. No subscription. No API key.
Plug in an NVIDIA GPU and it goes 3–15× faster — automatically. Linux + Windows. One click.

Developer Toolbox

Tools you reach for daily — built right into the app.

SHA-1/256/384/512 hash checker
CSS & JS minifier with size stats
JSON validator + pretty-printer

Just say it.

"Check the weather today." Searches the web, speaks the answer aloud.
"Schedule Pedro Tuesday 3pm." Books the visual calendar, WhatsApps Pedro to confirm.
"Build me a Flask dashboard." Writes it, tests it, iterates. Inspector validates before deploy.

LMIM v2 Tezcat — Developer Toolbox interface

LMIM v2 Tezcat — CUDA acceleration dashboard

CUDA Lightning

Auto-detects your NVIDIA GPU. 3–15× faster than CPU. No config.

llama.cpp + whisper.cpp on CUDA 12.x
RTX 20xx · 30xx · 40xx (compute 7.5+)
Falls back to CPU if no GPU detected

Token generation · Qwen 3.5 2B tok/s

CPU only 17

GTX 1650 Ti 80

RTX 3060 120

measured · same model · same prompt

Out of the box.

Qwen 3.5 bundled. Open the installer — model already inside. No download dance.
~120 tok/s on RTX 3060. ~80 tok/s on a GTX 1650 Ti. CPU fallback automatic.
Cloud APIs are optional. OpenAI, Anthropic, Groq — toggle in settings, never required.

Download v2 — Linux + Windows Download v2 manual (.md)

🐧 Linux · Live 🪟 Windows · Live v2.1 · In development

What LMIM Does

One AI. Infinite possibilities. All tools built-in.

🎙

Voice Control

Talk to LMIM. Hear it talk back. Fully on-device — no cloud, no API key, no audio leaving your machine.

✅ Whisper.cpp STT — speak any prompt
✅ Piper TTS — male voice replies
✅ Mic device selector
✅ EN + ES voices included

🐧

Cross-Platform

One-click installers. Linux AppImage + Windows .exe. Bundled model. No dependencies to hunt down.

✅ Linux: Ubuntu, Fedora, PopOS, Debian, Arch
✅ Windows: 10, 11 — InnoSetup installer
✅ CUDA on both platforms — auto-detection
✅ Same features, same model, same zero cost

🤖

Autonomous Workflows

Run agents that execute tasks across tools without constant input.

✅ Multi-step automation
✅ Tool orchestration
✅ Zero manual intervention

🔒

Local AI Runtime

Everything runs on your machine. No cloud dependency.

✅ Qwen 3.5 2B bundled via llama.cpp
✅ Optional cloud APIs (ChatGPT/Claude/Groq)
✅ Your data stays local

💬

WhatsApp from GUI

Manage WhatsApp directly inside LMIM — scan QR, send/receive, AI replies automatically.

✅ QR scan inside the app
✅ Auto-reply workflows
✅ Auto-reconnects on session expiry

💻

Code & Build

Generate scripts, apps, and automations directly from the interface.

✅ Auto-generate files
✅ Debug & fix errors
✅ Full project scaffolding

Release Log

v2 shipped. v2.1 is building. Always improving.

lmim-os — release log v2.1 preview · in dev

v2.1 NEXT ⬡ v2.1 in active development — three headline features dropping soon.

soon FEAT 🎨 Image generation — on-device, no API key, no upload. Describe it, see it.

soon FEAT 📄 Document mode — drop a PDF/DOCX, LMIM reads, edits, and answers. Local only.

soon FEAT 🪶 CPU-light mode — lean inference for no-GPU machines. Same privacy, smaller footprint.

May 05NEW🪟 WINDOWS — one-click Inno installer ships. Same features as Linux.

May 05NEWWindows CUDA — llama.cpp + whisper.cpp on GPU. Auto-detection.

May 05NEWWindows voice — Whisper STT + Piper TTS bundled in installer.

Apr 24v2.0⬢ TEZCAT — major release on Linux. Bundled model + CUDA + toolbox.

Apr 24NEWCUDA acceleration — llama.cpp + whisper.cpp on GPU. 3–15× faster.

Apr 24NEWDeveloper toolbox — SHA hash, CSS/JS minify, JSON validator, all in-app.

Apr 24NEWCalendar agent — "schedule X tue 3pm" books + WhatsApps confirmation.

Apr 24NEWService Switch — start/stop WhatsApp, Telegram, Email, Slack, Discord from GUI.

Apr 24NEWCampaign Blaster — bulk WhatsApp + email campaigns from CSV lists.

Apr 24IMPRAuto GPU detection — toggles CUDA on if compute capability ≥ 7.5.

Apr 24IMPRBundled Qwen 3.5 in AppImage (0.8B / 2B / 4B / 9B Q5 + 14B-A3B MoE).

Apr 24IMPRWhisper STT on GPU — ~0.3s transcription on RTX 3060.

Apr 24IMPREN/ES/PT/FR/DE TTS via Piper — five languages bundled.

Apr 15NEWVoice live on Windows — Whisper + Piper bundled in installer.

Apr 15NEWEN/ES language pill in chat bar — switch voice language on the fly.

Apr 09FIXWhatsApp QR auto-regenerates on session expiry — no manual restart.

Apr 09FIXStale auth files wiped on logout → fresh QR in ~3 seconds.

Apr 01IMPRCross-distro AppImage — built against glibc 2.35 (Ubuntu/Fedora/PopOS/Arch).

Apr 01IMPRHybrid inference — local Qwen default + optional ChatGPT/Claude/Groq.

Apr 01IMPRZero-setup install — chmod +x and run.

↓ Download v2 — Linux + Windows, ready now

What's Next?

v2 is live. v2.1 is already in the oven — here's a taste.

🪟

Windows v2

One-click InnoSetup installer. Bundled model. CUDA acceleration. All v2 features ready for Windows 10/11.

✅ All v2 features included
✅ CUDA acceleration on Windows
✅ Voice + WhatsApp + toolbox

✅ Live now

🍎

macOS Support

Apple Silicon + Intel. Same one-file philosophy. No Homebrew, no Xcode required. Metal acceleration on M-series chips.

✅ Apple Silicon (M1/M2/M3/M4)
✅ Intel Mac fallback
✅ Metal GPU acceleration

🔄 Coming soon

✦

Coming next

v2.1 — What's Coming

Three headline capabilities landing in the next drop — locally, privately, no cloud required.

🎨

Image Generation Generate images on-device — no API key, no upload, no waiting. Describe it, see it.

📄

Work on a Specific Doc Drop a PDF, DOCX, or TXT — LMIM reads it, edits it, answers questions about it. Your files stay local.

🪶

CPU-Light Mode A lean inference profile for machines without a GPU. Smaller footprint, same local privacy.

🛠 In active development

Ready to Get Started?

Download LMIM OS v2 'Tezcat' — bundled model, CUDA acceleration, developer toolbox. Linux + Windows. Free forever.

700+Downloads

20Countries

∞Possibilities

🐧 Linux AppImage — v2 'Tezcat' x86_64 · ~2.6 GB · v2.0 · Model + CUDA bundled 🪟 Windows Installer — v2 'Tezcat' .exe · v2.0 · Model + CUDA bundled · Windows 10/11

🐧 Linux: chmod +x LMIM_OS_2.0.0_TEZCAT.AppImage && ./LMIM_OS_2.0.0_TEZCAT.AppImage
🪟 Windows: Download LMIM_OS_TEZCAT.exe, double-click, follow the wizard.

Install via terminal:

🐧 Linux:

wget https://lmim.tech/downloads/LMIM_OS_2.0.0_TEZCAT.AppImage && chmod +x LMIM_OS_2.0.0_TEZCAT.AppImage && ./LMIM_OS_2.0.0_TEZCAT.AppImage

🪟 Windows: curl -L -o LMIM_OS_TEZCAT.exe https://lmim.tech/downloads/LMIM_OS_TEZCAT.exe && start LMIM_OS_TEZCAT.exe

Download v2 manual (README.md)

Other options:

v2 spotlight · Features · Changelog · Contact

🐧 Linux AppImage + 🪟 Windows Installer · Model + CUDA bundled on both · 📄 MIT License · No telemetry

Built with LMIM

Local AI in the Real World

Not demos. Not prototypes. Production deployments running LMIM's stack where privacy and offline reliability aren't optional.

🦊

Happy Fox App

Education · Mexico

A full English learning platform for kids — AI pronunciation scoring, on-device tutoring, and 155 weeks of curriculum. Runs entirely on the teacher's machine over LAN. Zero cloud.

🎙 Whisper STT — real-time pronunciation scoring

🔊 Piper TTS — instant audio for every word

🤖 Qwen LLM — Foxy AI tutor, on-device

🔒 100% offline — LAN only, no data leaves the school

#EdTech #LocalAI #Privacy #Offline

stack: llama.cpp · whisper.cpp · piper · sqlite · electron

Build yours with Enterprise →

Building something with LMIM? Tell us about it →

LMIM OS v2 — It just works.

Local AI That
Actually Does Things