Lucy AI Studio

A homefor your AI.

An always-on computer for AI agents, local models, and private storage.

See configurations Contact us on WhatsApp (opens in a new tab)
A warm-lit silver laptop with a photo editor on screen, a hand on the trackpad. Lucy’s machine running by itself in a dark bedroom while its owner sleeps.

01 / Your desk

Your computer
stays yours.

Give long-running tasks a machine of their own. Your laptop stays free for the work in front of you, and Lucy keeps your models, agents, and files together.

02 / Lucy

Lucy holds
the work.

Renders, agent runs, downloads. They keep moving on Lucy’s own compute while your laptop sleeps — pick the result up when you’re back.

While you sleep.

Start a model task, agent workflow, or creative render before bed. Lucy keeps working on its own compute while you rest.

Step away.

Leave your desk without keeping your laptop tied to a task. Your files, tools, and ongoing work stay together on Lucy at home.

Be here.

Stay with the people in front of you. Lucy handles background tasks on a separate machine, ready for you to pick up where you left off.

Promotional cinematic sequence showing generated visual worlds and title cards

Generate locally.Make it move.

Turn prompts and reference images into images and video with local creative workflows. The film shown is a promotional sample, not a recorded Lucy output.

From promptto output.

Work with local models, delegate tasks to agents, and keep the images, videos, and files you create in one workspace.

Your work. Within reach.

Send tasks from your phone.

Start a supported agent task and check its progress while Lucy handles the work at home.

Keep your files together.

Browse generated images, videos, and project files in a local library, without scattering them across devices.

Create locally.Then step away.

Build an image or video workflow, then leave the render running on Lucy. Your everyday computer stays free while the work continues.

Keep your library.Close at hand.

Keep generated media, project files, and shared content on your own storage. Access your library from devices on your home network.

In the press.

Thousands of glowing points form a floating spherical lattice, half of it still dissolving into drifting dust

How-To Geek

Review

“I think the Cephalon Lucy AI Studio Max is a fantastic piece of hardware. It’s far more capable than I was expecting it to be.”

Read the full feature
A ribbon of millions of glowing particles twisting through dark space, resolving from scattered dust into a flowing shape

xda-developers

Sponsored feature

“The Cephalon Lucy AI Studio isn’t just aimed at AI tinkerers and early adopters. This is a powerful computer system for people who want to do real work for real use cases. It’s designed to empower creators, developers, and enthusiasts who want to make AI a useful part of daily life.”

Read the full feature

Built to keep working.

Front-panel ports, high-speed networking, and five drive bays bring compute and storage together in one compact machine.

Twin-fan cooling.

Fans and heatpipes above the motherboard.

Lucy AI Studio

Lucy’s front connection row above the diagonal grille

Within reach.

USB-A · USB-C · 3.5 mm audio

Stay connected.

Pro: dual 2.5 GbE
Max / Ultra: 10 GbE

Ports vary by model. Reference unit shown.

Room to expand.

Five 3.5-inch SATA bays · Up to 100 TBSeparate NVMe system SSD · M.2 up to 8 TB

Drives sold separately.

Illustration of chip and memory pathways

Memory for more.

192 GBUltra

128 GBMax

32 GBPro

Confirm specifications with our team. Animation is illustrative, not a benchmark.

Storage Lucy / AI Studio

Keep your library close.

Give your projects room to grow, from source
files and documents to generated images and
video. Share them across your local network.

Drive bays
Five SATA bays
Expansion
Up to 100 TB
File sharing
SMB

Lucy AIOS

Built for local AI.

Built around Omarchy Linux, Lucy AIOS brings local models, agents, files, and creative tools into one environment. Choose what runs on your machine and give your agents access to the tools they need.

Think bigger.
Run it here.

Chat, draft, and work through documents with local models on dedicated compute. Larger memory configurations give you more room for demanding models and workloads.

Lucy AI Studio MaxRyzen AI Max+ 395
128 GB unified memory

Your agents.
Your command.

Start with preinstalled Hermes. Give it a goal and authorize the files and tools it needs. Add Claude Code, Codex, or OpenClaw, using compatible local models or supported cloud services.

  • Hermes
  • Claude Code
  • Codex
  • OpenClaw

Your command

Hermes waiting running finished
> draft release notes

and flag the open QA issue

Agent

Hermes Idle Receiving Working Done

Reads files, drafts notes, flags issues

  • Local model Local endpoint
  • Project files Granted access
  • QA check Open issue

Standing by Reading project files Running the local model Checking the open QA issue 3 steps complete

Output

Waiting for output

release-notes.md
qa-issue.md 1 blocker
Type Hand off Work Deliver

Imagine it.
Make it local.

Start with preinstalled ComfyUI, selected models, and prepared workflows. Turn a prompt or reference image into images and video, then refine the result on compute you control.

ComfyUI workflow · text-to-image flux1-dev.safetensors local · 1 device Queue
Load Checkpoint
modelflux1-dev.safetensors
CLIP Text Encode
positive
whale above a golden cloud sea
CLIP Text Encode
negative
blurry, low quality, watermark
Empty Latent Image
width1024
height1024
KSampler
steps24
cfg4.5
samplereuler
The image resolving inside the sampler
VAE Decode
sampleslatent
Save Image
filename_prefiximage_00001_
The finished image saved by the workflow: a whale gliding above a golden cloud sea

image_00001_.png · 1024 × 1024 · saved on this machine

Ready · 7 nodes · 1 prompt Queueing prompt · loading checkpoint Sampling · step 24 of 24 · 3.1 s/it Decoding · writing image to disk Done · image saved locally · no cloud round-trip flux1-dev · 126 TOPS · on device

Make it yours.

Connect power and Ethernet. Open your Setup Card’s tool address from a computer on the same network.

Ask a local model.

Runs on this machine. Nothing leaves it.

  1. Open WebUI.
  2. Pick a model, write a prompt.
  3. Review and save the answer.
Open WebUI picking a local Llama 3 model, sending a prompt, and copying the answer

Give Hermes a task.

Works only inside the access you grant.

  1. Choose a compatible model.
  2. Grant file and tool access.
  3. Send, then review the output.
Hermes choosing a compatible model, granting folder and tool access, and producing a report

Create with ComfyUI.

Rendered on this machine, not in the cloud.

  1. Open a workflow.
  2. Write a prompt or add a reference.
  3. Run, preview, save.
ComfyUI opening a Text to Image workflow, running a prompt, and saving the result

Find your Lucy

Choose the right memory and connectivity for your work.

Lucy Max selected. Contact us for configuration details.

Lucy Pro

Agents, workflows, and shared files.

Memory

32 GB

  • CPUAMD Ryzen 9 7940HS
  • GPURadeon 780M12 CUs
  • GPU memoryNot specified
  • AI performanceUp to 10 TOPS
  • SSD256 GB NVMe
  • NetworkDual 2.5 GbE
Shop Pro on Kickstarter (opens in a new tab)

Lucy Max

Local models and creative work.

Memory

128 GB

  • CPURyzen AI Max+ 395
  • GPURadeon 8060S40 CUs
  • GPU memoryUp to 96 GB
  • AI performanceUp to 126 TOPS
  • SSD1 TB NVMe
  • Network10 GbE
Shop Max on Kickstarter (opens in a new tab)

Lucy Ultra

Largest memory for local models.

Memory

192 GB

  • CPURyzen AI Max+ Pro 495
  • GPURadeon 8065S40 CUs
  • GPU memoryUp to 160 GB
  • AI performanceUp to 131 TOPS
  • SSD1 TB NVMe
  • Network10 GbE
Shop Ultra on Kickstarter (opens in a new tab)

Questions?

Find answers about Lucy AIOS, local workflows, storage, and support.

01What is Lucy AIOS?

An Omarchy-based Linux environment for local models, agents, files, and creative tools.

02What can run locally?

Models, agents, image and video workflows, and file storage. Compatibility and performance depend on the model, workflow, and hardware configuration.

03Does everything run locally?

No. Local workflows can process data on the device. Cloud models and online integrations may send data externally and require accounts or fees.

04Is 100 TB of storage included?

No. Five SATA bays support up to 100 TB of expansion. Drives are sold separately; usable capacity depends on the drives, formatting, and redundancy.

05What warranty support is available?

A one-year warranty covers manufacturing defects and functional issues under normal use, subject to applicable consumer law. Contact our team for terms and support.