Logen

Logen Desktop Studio Self-Healing Release

ComfyUI & Ollama power.
Zero terminal friction.

Logen is a native desktop studio for Windows and Linux x64 that runs FLUX.1-schnell, Stable Diffusion, and upscaling pipelines directly on your GPU—complete with project-linked AI chat threads, a 2-column workflow studio, and a 1:1 pixel lightbox.

Logen — Studio Inspector
Local GPU Active
Logen local FLUX.1-schnell render of a miniature ceramic robot in a sunlit terrarium 1:1 Full-Res Lightbox Ready
ACTIVE PROJECT Robot Terrarium 14 local renders · Thread #01
CHECKPOINT & WORKFLOW FLUX.1-schnell Q4 flux_schnell_txt2img · 4 steps
HARDWARE TELEMETRY Fits GPU (6 GB+ Listed) Validated on RTX A2000 12 GB
Ruby teapot render Cobalt teapot render Cyan glass orb render
01 / WORKSPACE ARCHITECTURE

Four integrated workspaces. Built to keep you in creative flow.

Switch between natural-language project chat, granular parameter tuning, project-organized gallery inspection, and hardware-aware model management inside one cohesive desktop window.

WORKSPACE 01 // MULTI-THREAD CHAT

Every conversation is a persistent creative project.

Click + New Chat at any time to start a fresh visual brief without losing earlier work. Each thread automatically names and tags every generated image with its project_id and project_name.

  • Auto-expanding multi-line composer: Paste complex prompts (Shift+Enter for new lines) and pick aspect ratios (1:1, 16:9, 9:16) inline.
  • Zero lost prompts on error: If a generation is interrupted or VRAM is busy, Logen preserves your exact prompt text in the composer for instant retry.
  • Rename & switch threads live: Renaming a chat thread automatically updates its folder name across your entire Gallery.
Active Project: Robot Terrarium + New Chat
USER PROMPT

“A small ceramic robot tending a lush miniature garden on a sunlit oak desk, macro lens, warm morning window light.”

Ceramic robot render inside Logen Chat
FLUX.1-schnell Q4 · 1024×1024 [Click for 1:1 Lightbox]
WORKSPACE 02 // TWO-COLUMN WORKFLOW STUDIO

Direct parameter control without node-graph clutter.

When you want deterministic control over a render, switch to the Workflow Studio. Human-readable parameter controls sit on the left while a sticky live preview canvas stays locked on the right.

  • Bundled production workflows: Text-to-Image (flux_schnell_txt2img, sd35_txt2img), Image-to-Image (sd15_img2img), and 4x RealESRGAN upscaling.
  • Inline parameter presets: Dial in Steps, CFG guidance, Sampler, Scheduler, Seed, and dimensions—then save reusable presets for your team or client style.
  • Custom ComfyUI JSON & Batch Queue: Import compatible ComfyUI API workflows or run multi-seed variations in Batch mode.
Workflow: Image-to-Image Colorway Seed: 904118 [Locked]
Steps20
CFG Guidance7.0
Denoise Strength0.62
Sampler / Schedulereuler / normal
Input ruby teapot Source Render
Output cobalt teapot Edited Output
WORKSPACE 03 // PROJECT GALLERY & 1:1 LIGHTBOX

Organized by project folder. Inspectable down to the pixel.

Stop hunting through numbered output filenames. Logen’s Gallery groups every generation under its project cover card with instant prompt search and a full-screen darkroom lightbox.

  • True 1:1 Full Resolution toggle: Inspect fine detail and upscaled edges at 100%–400% zoom or toggle Fit-to-Screen.
  • Keyboard-driven review: Use ← and → arrow keys to step rapidly through a project’s variations and Esc to return.
  • One-click roundtrip: Copy the image to clipboard, Save As, reveal the file on disk, copy the prompt, or click Open Project Chat to resume iterating.
Logen Gallery view
[1:1 Full Res] [← Prev / Next →] [Copy Prompt] [Open Project Chat]
WORKSPACE 04 // SELF-HEALING DIAGNOSTICS & MODELS

Self-healing local services & VRAM-verified downloads.

Logen eliminates the #1 pain point of local AI: broken background servers. If Ollama is asleep when you send a chat message, Logen spawns it silently in the background. If a model is missing or VRAM overflows, actionable recovery buttons fix it in place.

  • Actionable error taxonomy: Structured recovery for ollama_offline, ollama_model_missing, comfyui_offline, and vram_oom.
  • Hardware-aware Model Manager: Displays live GPU VRAM fit badges (Fits your GPU) alongside model license terms before you download.
  • Seamless desktop auto-updates: Built-in updater delivers new releases directly from titanxt.io/downloads/.
Model Manager · Hardware Fit GPU Detected
FLUX.1-schnell Q4 6 GB VRAM · Default · Apache-2.0
SD 1.5 (Checkpoints) 4–6 GB VRAM · Img2Img · CreativeML
FLUX.2 [klein] 4B Optional · Apache-2.0
Z-Image-Turbo Optional · Apache-2.0

Smart Setup & Self-Healing Engine

Detects your GPU and VRAM, installs ComfyUI and FLUX.1-schnell on first launch, auto-starts Ollama when needed, and delivers background app updates automatically.

Project-Linked AI Chat

Start a new project thread at any time, paste multi-line prompts, pick aspect ratios, and switch or rename project threads without losing earlier conversations.

Hardware-Aware Model Manager

Start with FLUX.1-schnell and SD 1.5. Browse eligible optional models including FLUX.2 [klein] and Z-Image-Turbo with GPU VRAM fit badges; each model has its own license and hardware requirements.

Project Gallery & 1:1 Lightbox

Every generated image is organized under its chat project with full metadata. Click any image to inspect at 1:1 full pixel resolution, navigate with arrow keys, or jump back to its chat thread.

Two-Column Workflow Studio

Fine-tune steps, CFG guidance, sampler, scheduler, seed, and resolution alongside a sticky live preview canvas, and save reusable parameter presets inline.

ComfyUI Workflows & Batching

Run bundled text-to-image, image-to-image, and upscaling workflows, import compatible ComfyUI workflow JSON, or generate multi-seed variations in Batch mode.

02 / HARDWARE FIT & COST SIMULATOR

Check what your GPU runs—and what you save vs. cloud plans.

Select your graphics card’s video memory (VRAM) below to see how Logen configures models for your tier.

SELECTED PROFILE

12 GB VRAM — Documented Validation Profile (NVIDIA RTX A2000 12 GB)

Logen was benchmarked and verified on an NVIDIA RTX A2000 12 GB. Runs FLUX.1-schnell Q4_K_S default text-to-image, SD 1.5 image-to-image, FLUX.2 [klein] 4B, Z-Image-Turbo, and 4x RealESRGAN upscaling smoothly.

DEFAULT STARTER MODEL FLUX.1-schnell Q4 (Listed 6 GB)
1-YEAR COST VS £25/MO CLOUD £0 Free · £49 Once (Save £251/yr)
03 / STRAIGHTFORWARD LICENSING

Start free forever. Upgrade once for £49 when you need batch & 4K.

No monthly retainers, no expiring credits, and a 30-day no-questions-asked money-back guarantee on Logen Pro.

STARTER STUDIO

Logen Free

£0 forever

Complete local AI image generation with our self-healing ComfyUI & Ollama runtime.

  • ✓ Local AI generation on your own GPU
  • ✓ FLUX.1-schnell default + SD 1.5 workflows
  • ✓ Project-linked multi-thread AI Chat
  • ✓ Project Gallery & 1:1 full-resolution Lightbox
  • ✓ Self-healing Ollama & ComfyUI diagnostics
  • ✓ Automatic background app updates
Download Logen Free
ONE-TIME LICENSE · 30-DAY GUARANTEE
PRODUCTION UNLOCKED

Logen Pro

£49 one-time

Unlock high-volume batching, 4K upscaling pipelines, and custom workflow imports.

  • ✓ Everything in Logen Free
  • ✓ Batch generation & multi-seed variations
  • ✓ 4K output resolution & RealESRGAN upscaling
  • ✓ Import custom ComfyUI API workflow JSON
  • ✓ Additional eligible models (third-party licenses vary)
  • ✓ Priority email support & lifetime updates
Get Logen Pro (£49 One-Time)

Pro licenses Logen application features, not third-party model rights. Internet required for Pro activation and periodic license checks.

04 / DEPLOYMENT & SYSTEM SPECIFICATIONS

From installer to first local render in three steps.

01

Install Logen

Download the Windows .exe or Linux .AppImage / .deb installer. The setup wizard detects your GPU and prepares the isolated Python and ComfyUI environment automatically.

02

Select Your Model

Start immediately with FLUX.1-schnell Q4 (Apache-2.0), or open Model Manager to inspect VRAM fit badges for SD 1.5, FLUX.2 [klein] 4B, and Z-Image-Turbo.

03

Create & Inspect at 1:1

Prompt in a project chat thread or tune parameters in the 2-column studio. Every output lands in your local project folder ready for 1:1 pixel inspection.

System Requirements & Platform Notes

All Installers & Checksums
Operating Systems
Windows 10/11 x64 (Logen-Setup.exe) and Linux x64 (Logen.AppImage, Logen-amd64.deb). macOS is not available in this release.
GPU & VRAM
NVIDIA GPU recommended. FLUX.1-schnell Q4 lists 6 GB VRAM; optional models require different memory footprints. Documented release validation performed on NVIDIA RTX A2000 (12 GB VRAM). Explore the Local AI GPU & VRAM Compatibility Index or the Local AI & ComfyUI Error Troubleshooting Reference.
System Memory & Disk
16 GB system RAM recommended. At least 30 GB free disk space required for environment setup, ComfyUI backend, and default starter model weights.
Model Licensing
FLUX.1-schnell, FLUX.2 [klein] 4B, and Z-Image-Turbo are distributed under Apache-2.0. SD 3.5 has separate stability.ai community/revenue terms. FLUX.1-dev and Qwen-Image-2.1 are disabled in this commercial build.
05 / DIRECT ANSWERS & TECHNICAL FAQ

Common questions on local AI generation, hardware, and licensing.

Straight answers on how Logen 1.4.0 runs FLUX.1-schnell and ComfyUI pipelines on Windows and Linux x64, what GPU memory you need, and how one-time licensing works.

What is the best local AI image generator for Windows and Linux without ComfyUI nodes?

Logen is a desktop app for Windows 10/11 x64 and Linux x64 that runs ComfyUI and Ollama in the background, so you never have to wire node graphs or debug Python environments by hand. You work in a clean window with project-linked chat threads, a two-column parameter studio, and a 1:1 pixel lightbox while your own GPU does the rendering.

What is the easiest local FLUX desktop app, and what GPU VRAM do I need?

Logen’s first-run setup wizard installs ComfyUI and downloads FLUX.1-schnell (Q4_K) automatically. You will want an NVIDIA GPU with at least 6 GB of VRAM for FLUX.1-schnell Q4_K (4 GB+ for SD 1.5), 16 GB of system RAM, and 30 GB of free disk space. We validated and benchmarked Logen 1.4.0 on an NVIDIA RTX A2000 12 GB.

Is there a local Midjourney alternative with a one-time purchase?

Yes. Logen Free is £0 forever and lets you generate with FLUX.1-schnell and Stable Diffusion 1.5 on your own hardware without monthly retainers or fast-hour queues. When you need batch generation (1 to 16 variations), 4K RealESRGAN upscaling, custom ComfyUI API JSON imports, and models like FLUX.2 [klein] 4B and Z-Image-Turbo, Logen Pro is a £49 one-time purchase with a 30-day refund guarantee.

Which operating systems does Logen 1.4.0 run on?

Logen 1.4.0 runs on 64-bit Windows 10/11 (Logen-Setup.exe) and 64-bit Linux (Logen.AppImage and Debian/Ubuntu Logen-amd64.deb). macOS is not supported.

Does Logen run offline, and what happens if a local service crashes?

After initial setup and model downloads, all rendering happens locally on your GPU and files stay in Documents/Logen/. If the local Ollama service stops between runs, Logen restarts it in the background and keeps your prompt text intact. Logen Free works offline after setup; Logen Pro needs an internet connection to activate your key and checks license status periodically, with a 24-hour offline window before reverting to Free.

Can I use Logen renders for commercial client work?

Logen Pro licenses the application features, while your rights to use generated images come from the specific model checkpoint you run. FLUX.1-schnell, FLUX.2 [klein] 4B, and Z-Image-Turbo are released under the Apache-2.0 license, which permits commercial use subject to Apache-2.0 terms. SD 3.5 carries separate Stability AI revenue conditions, and non-commercial weights (FLUX.1-dev and Qwen-Image-2.1) are disabled in this build.

LOGEN DESKTOP STUDIO — AVAILABLE NOW

Ready to run your own AI image studio?

Free download for Windows and Linux x64. No credit card required to install and generate.

FREE 11-PAGE PDF HANDBOOK + LOCAL AI FIELD NOTES

Get the Free 11-Page Local AI Image Generation Handbook

See what GPU VRAM you need for FLUX.1-schnell and SD 1.5, how checkpoints and LoRAs work, and five repeatable workflows on Windows and Linux x64. Enter your email to open the 11-page PDF right away and get our six-part weekly local AI field notes.

6–24 GB VRAM Tables Cloud vs. Local Math 5 Studio Workflows
How signup uses your email

We register your email with the TitanXT newsletter on beehiiv to send the 11-page PDF link, our six-part weekly local AI sequence, and occasional Logen updates. You can unsubscribe in one click from any email or by writing to support@titanxt.io. Read beehiiv's privacy policy (new tab).

Scroll to Top