Back this project →
Kickstarter campaign · Pre-launch

Local AI that actually works.
No subscriptions. Ever.

Run 150+ AI workflows on your own machine. Images, video, music, text, 3D. Fully offline. Buy once — yours forever.

Back this project on Kickstarter →

First 150 backers: £49  ·  After that: £89  ·  No subscription, ever

Windows + Mac
Desktop app
150+
AI workflows
Local
Private execution
One-Time
No monthly fees
Real outputs from Promptus

What you can create — locally, privately, instantly

Every image and video below was generated on a local machine using Promptus. No cloud. No billing meter. No watermarks.

Promptus CosyFlows local AI generation
Local AI
AI generation — runs on your machine
Flux.1, SDXL, SD 3.5 · no cloud · no billing
Hunyuan 3D generation
3D Generations
Hunyuan & Blender — local mesh output
No cloud render farm needed
Wan 2.1 video frame
Video
Wan, LTX Studio — Local generation
Text-to-video · Image-to-video
Flux.1 image generation
Images
Stable Diffusion, Flux, Qwen, +more
Optimized for local · runs on RTX 3060+
AI music generation
Music
AI music — generated locally
No Suno subscription required
Z-Image Turbo workflow
CosyFlows
Z-Image Turbo
Lip sync workflow
CosyFlows
Lip Sync — InfiniteTalk
Image to video workflow
CosyFlows
Image-to-Video
Enhance Prompt feature
Playground
✨ Magic Promptus
The app itself

Create anything. No limits. No subscriptions.

From visual workflow editing to one-click generation — every surface is designed so you never touch a command line.

Promptus — Workflow Editor
ComfyUI node graph editor
Visual workflow editor — no terminal required
Promptus maps any ComfyUI JSON into a clean graph. Dependencies resolve automatically. One click to run.
Promptus — CosyFlows Library
CosyFlows workflow browser
150+ ready-to-run CosyFlows — browse and launch
Every workflow ships with example outputs. Pick one, adjust your prompt, and generate — no setup, no configuration.
Promptus — Playground · Enhance Prompt
Promptus Playground — Enhance Prompt feature
The Playground — Image / Video / Audio in one view
Switch modalities with a tab. Enhance Prompt rewrites your text using a local LLM before generation — better results, fully offline.
Image Video Audio ✨ More Magic
See it in action

From install to generation in under 10 minutes

Watch how ComfyUI workflows go from broken JSON to running output — no terminal, no Python.

Promptus — Local AI Demo
Promptus demo — local AI workflows
▶
Watch the demo · 2 min
Step 01
Install Promptus
Download for Windows or Mac. One-click installer, no Python, no terminal, no config files.
✓ Desktop ready
Step 02
Choose or import a workflow
Pick from 150+ included CosyFlows, or drag in your own ComfyUI JSON — it maps and resolves automatically.
✓ Less friction
Step 03
Generate locally
Run on your own hardware. Keep full control over prompts, files, and output. No cloud, no billing meter.
✓ Private by default
Why buy once matters

You're already paying £140 a month
for tools you don't own.

ChatGPT. Gemini. Runway. Midjourney. ComfyUI Cloud. That's £1,730 a year — and you don't own a single byte of what you've made. Promptus ends that.

What you're paying now
ChatGPT ChatGPTMonthly AI subscription — cloud only, no ownership.
£20/mo
Claude GeminiYour prompts processed on someone else's server.
£20/mo
RunwayML RunwayML£76/mo for video you rent, not own.
£76/mo
Midjourney Midjourney + othersPer-image billing, content filters, lock-in.
£24/mo
Monthly total
£140/mo
VS
Promptus — buy once
Own your AI
One-time purchase · No subscription

Promptus runs on your machine. Your prompts never leave your device. No monthly meter, no per-generation billing, no cloud lock-in. Pay once. Use forever.

See reward tiers below ↓
Head to head

Stop paying rent on tools you should own

Same output quality. Zero monthly bills. Plus 150+ workflows that run instantly.

BEST VALUE
Promptus		 RunwayML Pika Labs
Monthly cost £0 / moOne-time £49–£89 · yours forever £76 / month £55 / month
ComfyUI import ✓ Drag, drop, run ✗ ✗
150+ pre-built workflows ✓ CosyFlows included ✗ Build from scratch ✗
Fully offline ✓ No internet needed ✗ Cloud only ✗ Cloud only
Data privacy ✓ Never leaves your device ✗ Cloud processed ✗ Cloud processed
Content restrictions ✓ Your hardware, your rules ✗ Heavy filters ✗ Limited
Year-1 total cost £49 total. Done. £912 £660
System requirements

Will it run on your machine?

Promptus works at every hardware tier — from a GTX 1060 to an RTX 5090. Here's exactly what runs at each level.

Minimum
Nvidia GTX 1060 6GB+
or CPU-only (slower)
  • Qwen3 7B, Mistral 7B, Phi-4
  • Stable Diffusion 1.5, SDXL
  • Z-Image Turbo (2s generations)
  • Hunyuan 3D (fast)
  • CosyFlows lightweight
  • Wan 2.1 video — slow, short clips
★ Recommended
Nvidia RTX 3060 12GB+
or Apple M2 / M3 16GB
  • Qwen3 14B, Phi-4 14B (text + agents)
  • Flux.1 Dev (full quality images)
  • Stable Diffusion 3.5 Medium
  • Wan 2.1 video 720p — comfortable
  • Full CosyFlows library
  • DeepSeek-R1 32B (coding agents)
Full power
RTX 4070+ / RTX 5090
or Apple M3 Max 36GB+
  • Llama 3.3 70B — full quality, fast
  • Qwen3 30B, 72B variants
  • LTX-2 & LTX-2.3 video — smooth
  • Wan 2.1 1080p video — smooth
  • Flux.1 Dev large batch
  • Full multi-step OpenClaw agents
Mac
Apple M1+ (8GB unified)
macOS 12 Monterey or later
  • M1 8GB: basic image generation
  • M2/M3 16GB: recommended for Mac
  • M3 Max / M4 Pro: full video speed
  • All Apple Silicon supported
  • Intel Macs: CPU-only mode
Not sure if your machine qualifies? Drop your specs in the Kickstarter comments or message us — we'll tell you exactly what will run on your setup. Email us at support@promptus.ai.

No GPU? Promptus supports cloud-backed generation through API routing — slower, but it works while you wait for hardware. AMD GPU support is on the roadmap for Q3.
Kickstarter rewards

Own your AI workflow instead of renting it.

Lifetime licence. Choose your platform at redemption.

🔥 Best deal
Super Early Bird
£49
Retail £89
Save 45% — limited spots
  • Promptus App — Windows or Mac (you choose at redemption)
  • Lifetime licence — 50+ open-source models at launch
  • New model support as models release
  • 150+ CosyFlow workflows included
  • Your name in the credits
  • Private Founder Discord channel
  • Vote on the next 5 CosyFlows built
Back Super Early Bird →
Early Bird
£69
Retail £89
Save 22%
  • Promptus App — Windows or Mac (you choose at redemption)
  • Lifetime licence — 50+ open-source models at launch
  • New model support as models release
  • 150+ CosyFlow workflows included
  • Priority support channel
Back Early Bird →
Standard
£89
Retail price
  • Promptus App — Windows or Mac (you choose at redemption)
  • Lifetime licence — 50+ open-source models at launch
  • New model support as models release
  • 150+ CosyFlow workflows included
  • Priority support channel
Back at Standard →
Windows + Mac Bundle
£129
Combined £178
Save £49 — both platforms
  • Promptus App for Windows
  • Promptus App for Mac
  • Lifetime licence on both platforms
  • 50+ open-source models at launch
  • Future model updates included
  • Priority support channel
Back Windows + Mac Bundle →

What happens after you back us

Day 1 after funding closesYou receive an email confirmation with your backer ID.
Within 30–60 daysYou receive your download link and licence key by email.
At redemptionChoose Windows or Mac. Download the installer. Activate with your key. Done.
From the community

Built in public. Tested by real creators.

380 beta users across 251 Discord members have stress-tested this — and told us exactly when it breaks.

★★★★★
"I had 40 hours invested in a ComfyUI workflow for client work. A Python update broke it. I spent a week trying to fix the dependencies. I dragged the JSON into Promptus. It ran in four minutes."
EM
Eric Martin
Graphic Designer · Beta user
★★★★★
"I had 20 ComfyUI workflows sitting unused. I dragged them in, they all ran. I haven't opened a terminal since."
JR
Jose Romero
Illustrator · Beta tester
★★★★★
"Overall it's an impressive system that makes it easy to produce great results. Promptus has enabled me to create very impressive Pre-Raphaelite style images."
JH
John Howard
Photographer · Beta tester
FAQs

Your questions answered

Yes — pay once and own Promptus forever. No monthly fees, no annual renewal, no subscription traps. All future updates — every new model, every new workflow — are included. Because AI runs on your local hardware, there's nothing to bill you for. One payment, yours forever.
An Nvidia GTX 1060 6GB+ is the minimum for image workflows. An RTX 3060 12GB is the sweet spot for video generation. RTX 4070+ or Apple M3 gives you the full experience. Not sure about your machine? Drop your specs in the Kickstarter comments — we'll tell you exactly what runs.
Yes — Apple M1 and later are fully supported on macOS 12 Monterey or later. M2/M3 with 16GB unified memory is the recommended Mac configuration for video generation. Intel Macs are supported in CPU-only mode. You choose your platform (Windows or Mac) at redemption when the campaign closes.
Once the campaign closes, you'll receive an email confirmation with your backer ID. Your download link and licence key will arrive within 30–60 days. At that point you choose Windows or Mac, download the installer, activate with your key, and you're running. We'll keep you posted on progress in the backer updates.
The app runs fully locally on your machine — it doesn't require Promptus servers to function. Once installed, it works completely offline. You'd keep the version you have and continue using it indefinitely. We've been building in public for two years with 380 beta users and 251 Discord members — we're not going anywhere, but even if we did, your licence keeps working.
Yes — drag any ComfyUI JSON file in and Promptus maps the full node graph, resolves dependencies, and runs it automatically. No terminal, no manual package installs, no red node hell. Workflows that have been sitting broken for months run in minutes. It's one of the most-mentioned things in beta tester feedback.
Promptus has no built-in content filters. Your hardware, your rules, your output. Nothing is logged, cached, or sent to our servers. Local execution means no cloud policy dictates what you can create. You are responsible for complying with the terms of the underlying open-source models you use.
Yes — fully offline once models are downloaded. The initial model downloads require internet, but after that Promptus runs with no connection at all. Your prompts, outputs, and workflow files never touch a server. Promptus also supports optional API routing for frontier models (GPT-4o, Claude) when you want them, but that's entirely opt-in.
Get started

Local AI that actually works.
Yours forever.

150+ workflows. 50+ open-source models. Fully offline. No subscription. All future updates included. First 150 backers: £49.

Back this project on Kickstarter →
✓ Windows + Mac ✓ Instant download after campaign ✓ Fully offline ✓ Data never leaves your device ✓ No subscription, ever