Your local AI workstation.
lucebox.com Source★ 14MIT DiscordAI Models & Agents / Model Training & Inference Infrastructure
- Monthly visits
- 8.1K
- Up60.4%MoMSep 2026
- Public revenue
- Undisclosed
- No public revenue evidence yet
- Entry price
- $5999/mo
- Higher than 99% of peersNiche median $20/mo
- Launched
- 20d ago
- 2026-10-26 · Est.
Traffic
Marketing moves
Launches, search & content, referrals, promotions- Our agents are working 24/7 to improve lucebox performances 🚀 引用帖 @pupposandro: We built our own inference engine, and it ships on every @luceboxai That means Lucebox customers gets day 0 support for new models and weekly optimisations.
@davideciffa · 2026-11-12 · 6 likes · 1 reposts · 266 views
- Another day, another box of boards at @luceboxai
@pupposandro · 2026-11-11 · 84 likes · 4 reposts · 1,905 views
Lucebox introduces custom inference engine shipped on every deviceProduct Activity
Lucebox built its own inference engine designed specifically for its hardware (495 + R9700) and ships it on every Lucebox. The company stated this provides customers with day 0 support for new models and weekly optimizations.
Nov 12 · 3d agoBuild in PublicVs. IncumbentX Post
Lucebox runs its own CI on one LuceboxPromotions
Lucebox says one Lucebox now reviews every pull request on its main repository, tells why a job failed, and runs the Linux jobs moved to it from GitHub Actions. The code never leaves the office. Hosted CI bots cost about $1,340 a month; the box costs $8,199 once and about $310 a month to run, and pays for itself in about 7 months.
Nov 4 · 11d agoBuild in PublicX Post
Lucebox launched on UneedLaunch
Lucebox
Oct 26 · 20d agoSource
At the time: 8.1K visits/mo4 concurrent movesLaunched on Uneed (#7)Launch
Lucebox — The computer for local AI
Oct 26 · 20d agoLaunch page
At the time: 8.1K visits/mo4 concurrent moves
Third-party reviews & case dossiers
Acquisition breakdowns, engineer reviews & open-source metricsLucebox is a plug and play local AI computer for teams, with RTX 3090, Ryzen AI MAX+ 395, 128 GB unified memory, and tuned inference software.
Open Source & Agent Ecosystem
- GitHub Stars
- ★ 14
- Forks
- 3
- License
- MIT
- Primary language
- Go
Agent, MCP & Developer Integration
Agent Skill- Ecosystem & integrations:
- Claude CodeCodexOpenCodeHermesOpenClawOh My PiOpen WebUITailscaleClaude
Traffic sources
Traffic breakdown
Channels
- Direct33%
- Organic Search28%
- Referrals13%
- Social7%
- Other19%
Countries
- United States53%
- Türkiye12%
- Netherlands10%
- Taiwan10%
- India9%
Search
Domain authority —Domain registered 2016-10-31
| Keyword | Monthly volume |
|---|---|
| laguna xs 2.1 | 1,110 |
| lucebox | 540 |
| what is local inference | 210 |
| amd lucebox | 60 |
| deepseek v4 flash r9700 | — |
Traffic data: Vaneform
Pricing
Trial: None · 1 paid tier ($5999) · Observed 2026-11-04 · Source page
Entry price $5999/mo, Higher than 99% of peers; “Model Training & Inference Infrastructure” median is $20/mo.View pricing benchmark
Pages & changes
Key pages
Weekly snapshotsOfficial blog & changelog
Recent articles & product updatesLing 3.0 Flash on DGX Spark: Up to 141.9 tok/s with Adaptive DSpark and FlashKDA
We got AntLing’s 124B model running on one NVIDIA DGX Spark. The result is mixed: Lucebox reads prompts faster than llama.cpp, while normal token-by-token generation is about tied. Our speculative path reaches 141.9 tok/s in a favorable tes
Continuous batching: Qwen3.8-27B at 301 tok/s across five clients, up to 1.97× llama.cpp
Lucebox now serves several requests at once and refills each slot the moment it frees up. On one R9700, Qwen3.8-27B with DFlash2 goes from 106.6 tok/s at one client to 300.9 at five, up to 1.97× llama.cpp at equal concurrency. Here is how t
ROCm beats Vulkan on Strix Halo
A Strix Halo mainboard 283.9 tok/s prefill at 123K prompt tokens 38.0 tok/s speculative decode adaptive width, q5 chosen on every step +47% / +35% prefill / decode over llama.cpp Vulkan at 123K prompt tokens Single runs at temperature 0 wit
Inference-aware cooling: 29% less fan speed on the AMD R9700, same temperatures
A fan controller only knows the past: the heat that has already happened. The AI model running on the card knows the future: how many tokens are queued, how long the prompt is, whether an agent is about to call again. On the AMD Radeon AI P
Changes
No pricing, plan, or positioning changes detected since tracking started.
Design
Minimal Monochrome (Light canvas, airy density) featuring split copy & visual layout and photography / live video visuals.
First-screen preview · click to zoom
First-screen pitch & conversion flow
“Your local AI workstation.”
Split Copy & Visual hero structure paired with direct checkout / view pricing CTA and stacked avatars & testimonials.
Conversion path: Split Copy & Visual + Photography / Live Video → Direct Checkout / View Pricing · Trust bar: Stacked Avatars & Testimonials
- Category Positioning Anchor
- Metrics Strip
Palette · click to copy
- Typography
- 8.0
- Color & Contrast
- 8.0
- Layout & Spacing
- 9.0
- Craft & Polish
- 8.0
- Distinctiveness
- 7.0
Structure
- Fonts
- Inter
- Hero layout
- Split Copy & Visual
- Primary visual
- Photography / Live Video
- Conversion CTA
- Direct Checkout / View Pricing
- Social proof
- Stacked Avatars & Testimonials
- Navigation
- Full-Width Top Bar
- Theme & density
- Light · Airy
- Corners & depth
- Small Radius · Hairline borders
Preview full AI design brief (includes :root CSS variables & layout rules)Style tokens only; never copy brand names, logos, copy, or images
# Design Brief: Minimal Monochrome
Monochrome palette with a single accent color, hairline dividers, and generous whitespace.
## Design Tokens
```css
:root {
--background: #ffffff;
--surface: #f5f5f4;
--foreground: #111111;
--muted-foreground: #777777;
--accent: #f15a24;
--radius: 4px;
--font-heading: "Inter", ui-sans-serif, system-ui, sans-serif;
--font-body: "Inter", ui-sans-serif, system-ui, sans-serif;
}
```
Color area ratio: Background #FFFFFF (~88%) · Surface #F5F5F4 (~3%) · Text #111111 (~6%) · Muted Text #777777 (~2%) · Accent #F15A24 (~1%). Background and surface dominate the canvas; reserve the accent color for primary buttons and active states.
## Layout & Conversion Structure
- Hero layout: Split Copy & Visual
- Navigation: Full-Width Top Bar
- Theme & density: Light · Airy
- Border radius & elevation: Small Radius; no drop shadows—separate layers with hairline borders and whitespace
- Primary visual medium: Photography / Live Video
- Conversion CTA: Direct Checkout / View Pricing · Trust bar uses Stacked Avatars & Testimonials
- Headline hook: Category Positioning Anchor
- Hero components: Metrics Strip
- Spatial layout notes: Split Copy & Visual hero structure paired with direct checkout / view pricing CTA and stacked avatars & testimonials.
- Visual summary: Minimal Monochrome (Light canvas, airy density) featuring split copy & visual layout and photography / live video visuals.
## Avoid
- Multi-stop neon gradients
- Heavy drop shadows
- More than one competing accent color
Implement the UI following the design tokens and layout specifications above: use only these color variables, and reserve the accent color strictly for primary actions and active states. Adopt only this visual style—never copy any reference product’s brand name, logo, copy, or imagery.
Questions & evidence
- How much does Lucebox cost?
- Developer: $5,999/once. Observed 2026-11-04. Check the source for current billing terms, usage limits and taxes.Source
- How much website traffic does Lucebox get?
- Vaneform estimates 8.1K website visits in 2026-09. Visits can include repeat visits; this estimate measures web traffic separately from paying customers or revenue.Source
- What is Lucebox and what platforms does it support?
- Lucebox is catalogued under Model Training & Inference Infrastructure: Your local AI workstation.. Supported platforms: Web.
- Is Lucebox open source?
- Lucebox has a public repository (https://github.com/Luce-Org/lucebox) under the MIT license, recording 14 GitHub stars.Source
Related products in this category
These products share a catalog category. Compare customer tasks, features and deployment requirements before treating them as alternatives.
| Product | Entry price | Visits | Revenue |
|---|---|---|---|
| OllamaGiantMost popular way to build with open models. | $20/mo | 9.4M | Undisclosed |
| Groq®Groq was built for this. | — | 3.6M | Undisclosed |
| BytePlusAI Film Award 2027 Powered By BytePlus | — | 3M | Undisclosed |
| ModelScopeFast-forwarding AI development with open-source community. | — | 2.9M | Undisclosed |
| RunpodExperiment, train, fine-tune, deploy, scale on one platform | — | 2.8M | Undisclosed |
| LM StudioThe agent for your computer | $20/mo | 2.5M | Undisclosed |