LumaDock (@LumaDock) header
LumaDock (@LumaDock) profile picture

VPS, GPU and bare metal servers for developers and businesses. Deploy in seconds across 9 global locations, with unmetered traffic and no setup fees.

London, England
Joined March 2025
Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)Photo by LumaDock (@LumaDock)
Ever wonder who's behind "deploy in seconds"? Our infrastructure team. More capacity went live in London, Amsterdam and Madrid, and we completed network upgrades in Frankfurt and Bucharest. The work of making it faster never really stops. Thank you!
1
4
6
91
#Webmin edits your server's real config files from a browser, and there is no default password to look for. Our complete guide covers install, login, security and where it stops, tested on a fresh VPS. Deploy Webmin with 1 click! lumadock.com/tutorials/webmi…
1
1
39
Just been handed a SQL Server you didn’t set up? Our #MSSQL guide explains what you’re looking at and where T-SQL will surprise you if you came from PostgreSQL or MySQL. lumadock.com/tutorials/mssql…
1
22
Everything you need for #Miniconda on one page: install on Linux, Windows or Mac, environments that don’t break each other, PyTorch, updates, licensing and the usual errors. 12 deeper tutorials sit behind it when you need the full steps: lumadock.com/tutorials/minic…
1
16
Self-host #Rakazo on a VPS, the open-source Grok Bot alternative. One installer, Caddy for HTTPS, a signup allowlist before first start and your own model keys. Docker Engine 26 is the only hard requirement. lumadock.com/tutorials/self-…
1
1
29
#Jev, from @typesafeai, is an AI model with no text output. Fixed answers in, typed decision plus probabilities out, under 500 ms, $0.042 per million tokens. Ties GPT-5.6 Terra on their evals at 75x lower cost, and yes! It plays Doom. lumadock.com/blog/what-is-je…
1
1
72
Still self-hosting @n8n_io from a 2025 tutorial? Since 2.x: latest tag renamed, $env blocked in Code nodes, MySQL gone, and from 3.0 in October it's Docker only. The guide we keep #n8n current, sizing table and compose file included: lumadock.com/tutorials/self-…
1
1
58
Seeing Conny on the shelves still feels surreal. And now her story is almost ready to travel even further. 💙📖✨ Conny and the Stars from Another Sky is coming soon in English. #ConnyAndTheStars #ComingSoon #FantasyBooks #BookTwitter #EnglishEdition
2
6
63
Meta Muse Spark 1.3 is out and the weights are staying on Meta's servers. One Muse model does run on your own GPU though, and it fits in 24 GB with a single Ollama command. Also in the post: the tier that's 10x cheaper because Meta trains on your prompts. lumadock.com/blog/what-is-me…
1
33
More #VPS capacity, live in #Frankfurt. Our CEO spent the day at our partners @firstcolo's Werkhaus facility watching our expansion go in. Next to @DECIX on 100% renewable power. Room for your next project, deployed in seconds.
1
1
2
39
Self-hosting @grafana takes an afternoon and it's one of the more satisfying things you can put on a #VPS. Our full guide: install, first login, #Prometheus, dashboards, alerting, hardening. Written and verified against 13.2 lumadock.com/tutorials/grafa…
3
1
51
"18B active parameters" does not mean GLM-5.3-Flash runs on 18B-class hardware. The full 320B checkpoint wants ~386 GB of VRAM. Self-hosting starts at an 8-GPU node. The developer's guide, with verified benchmarks and real costs - link in comments
2
1
21
One installer, one command, a working #LLM on your own hardware. Our complete @ollama guide is live: install, hardware sizing, model picks, API use and server hosting, linking out to 15 deeper tutorials. lumadock.com/tutorials/ollam…
1
34
A mystery lab dropped a free 1M-context coding model on @OpenRouter and won't say who they are. The tokenizer may have already answered for them. Everything verified so far, plus the real score behind the viral 80% benchmark claim: lumadock.com/blog/what-is-ox…
1
85
We wrote 25 tutorials on @claudeai #ClaudeCode. Installs, MCP, subagents, hooks, tmux sessions that survive your wifi and the errors you're trying to fix right now. Took weeks but worth it. This tool changed how our team ships. Hats off to @AnthropicAI lumadock.com/tutorials/claud…
1
33