The AI at 127.0.0.1. Run a local LLM on your own machine. No $20–200/month subscription, no data leaving your computer, no cloud dependency.
$240–2400
ChatGPT / YEAR
VS
$99
127LLM · DISCOUNT (WAS $127)
✳ PAYS FOR ITSELF IN WEEKS ✳
The Math
ChatGPT Subscription
Year 1: $240–2400
Year 2: $240–2400
Year 3: $240–2400
Total: $720–7200
Own nothing.
127LLM
Year 1: $99
Year 2: $0
Year 3: $0
Total: $99
Own your AI.
You save up to $6,100+. And you own your stack.
✳ WHAT YOU GET ✳
127 Hub + CLI
Local web dashboard and CLI, pre-configured for your hardware. Run from the browser, your terminal, or both — whichever fits your workflow. Detects Apple Silicon, NVIDIA, or CPU and sets optimal flags automatically.
DETECTS · CLI · WEB · BOTH
5 Curated Models
Selected for speed and quality on consumer hardware. A classifier (0.8B), a fast chat model (3B), an instruct model (7B), a code expert, and an embedding model for local RAG.
VERIFIED ON 16GB SYSTEMS
Upgrade Pipeline
One-click version upgrades without re-downloading multi-GB models. The repackaging system swaps the runtime while preserving your weights.
DOWNLOAD ~30MB INSTEAD OF ~4GB
Cross-Platform
Runs on macOS, Windows, and Linux — Intel x86_64 and ARM64. Use CPU, GPU, or both simultaneously. Auto-detects your hardware and tunes flags for maximum performance out of the box.
MACOS · WINDOWS · LINUX · INTEL · ARM64
Privacy by Design
Every inference runs on your machine. No data ever leaves your computer. No account, no phone-home, no telemetry.
100% LOCAL · 0% PHONE-HOME
Universal Compatibility
Works with Hermes, OpenClaw, and any llama.cpp harness. Built on open-source GGUF — your models are portable across the entire ecosystem. No vendor lock-in.
HERMES · OPENCLAW · LLAMA.CPP · GGUF
✳ ROADMAP ✳
v0.1 — Pre-Alpha · NOW
Gauging interest. Landing page + waitlist signup. Your input shapes the product.
v0.2 — Alpha · Q4 2026
Early access for waitlist members. 127 Hub dashboard + CLI MVP, 3 curated models, auto-config for Apple Silicon. macOS + Linux.
v0.3 — Beta · Q1 2027
All 5 models, upgrade pipeline, NVIDIA GPU + Intel support, CPU/GPU compute mode, community recipes, Hermes & OpenClaw harness compatibility.
v1.0 — Launch · Q2 2027
One-time purchase available. Community subscription launches. Cross-device sync. Full Windows, Linux, macOS, Intel & ARM64 support. CPU + GPU modes.
✳ PRICING ✳
127LLM
$99 one-time · intro pricing
Discounted pre-launch price (will be $127 at launch). Everything you need to run local AI on your own machine. No subscription required.
127 Hub dashboard
5 curated models
Auto-configuration
Upgrade pipeline
30-day community access
Lifetime of the product
Community
$12/monthor$99/year
ANNUAL BILLING SAVES YOU 31% · OPTIONAL
For enthusiasts who want to go deeper. Monthly model packs, workshops, and a growing recipe library.
Everything in 127LLM
Monthly model pack (new models + benchmarks)
Prompt engineering workshops
Recipe library (RAG, agents, one-liners)
Priority support
Cross-device sync
$99 IS A PRE-LAUNCH DISCOUNT · REGULAR PRICE $127 AT LAUNCH · WAITLIST LOCKS IN THE $99 RATE
✳ JOIN THE WAITLIST ✳
Join the waitlist to lock in the $99 pre-launch discount ($127 at launch). Be the first to know when 127LLM ships.
SIGN UP FOR UPDATES
✳ SPREAD THE WORD ✳
Help others discover local AI. Pick a platform, grab some copy, and boost the signal. Every share reaches someone who needs this.
Personal AI. Offline. Forever. $99 one-time (intro pricing), run local LLMs on your machine. No subscription, no phone-home, no cloud. macOS, Windows, Linux, Intel, ARM64, CPU or GPU.
127LLM — the AI at 127.0.0.1
127llm.1abundant.org
For LinkedIn — Professional
I've been testing 127LLM — a personal AI that runs fully offline on your own machine. Pay once ($99 intro pricing, $127 at launch), own it forever.
Runs on macOS, Windows, and Linux — Intel x86_64 and ARM64, CPU, GPU, or both simultaneously. Use it from the browser, the terminal, or both. Works with existing tools like Hermes and OpenClaw.
No subscription. No data leaving your computer. No cloud dependency.
127llm.1abundant.org
For Reddit / Hacker News — Community
127LLM — Personal AI that runs fully local. $99 one-time (intro, will be $127 at launch), no subscription. Uses llamafile / GGUF under the hood, works with Hermes, OpenClaw, and any llama.cpp harness. Cross-platform: macOS, Windows, Linux, Intel, ARM64, CPU, GPU, or both. CLI and web interface included.
127llm.1abundant.org