xCloud joins LLM Hosting Pro with managed Ollama and LocalAI plans
We've added xCloud as a host, with two managed CPU plans for running open-weight models on Ollama or LocalAI. We've also added Microsoft's Phi family to the site, starting with Phi-4.
We've added xCloud to LLM Hosting. It brings two managed plans for running open-weight models on your own private server: xCloud Ollama Hosting and xCloud LocalAI Hosting. Both plans are now listed on the site, alongside a new model family, Phi.
About xCloud
xCloud started out focused on WordPress. It built its name as a managed hosting and server-management platform for WordPress sites, and today it hosts more than 162,000 sites across 31 server locations. The company, xCloud Hosting, LLC, was founded in 2023.
Since then it has grown into a wider "host anything, deploy anywhere" platform for WordPress, Laravel, Node.js, Docker and AI agents. You can use one of xCloud's managed Cloud VPS servers, or connect a server you already have with DigitalOcean, Vultr, Linode, AWS, Hetzner or on bare metal. LLM hosting is the newest part of that platform.
Because WordPress is where xCloud began, it is a good fit for any LLM project that needs to work with WordPress. You can run a WordPress site and a private model on the same platform, managed from one dashboard. That suits projects like an AI chatbot or support assistant on a WordPress site, content and SEO tools that call a model from WordPress plugins, or search over your own posts and documents. Your data stays on your own server instead of going to a third-party AI API.
The two plans
xCloud Ollama Hosting costs US$59.99 to US$84.99 a month. It comes in two Cloud VPS sizes with 16 GB or 24 GB of RAM, and runs quantised models of up to about 27B parameters. Ollama installs in one click, you manage models from the dashboard, and you can add Open WebUI or LibreChat as a chat interface.
xCloud LocalAI Hosting costs US$59.99 a month for a Cloud VPS with 16 GB of RAM, and runs models of up to about 14B parameters. LocalAI puts text, vision, voice and image models behind one OpenAI-compatible API, with user accounts, per-user API keys, and agents that can use tools, RAG and MCP.
Both plans are CPU-only, with an OpenAI-compatible API, free SSL, and backups. If you need larger models, xCloud also lets you connect your own server, including GPU machines. Your first server is free, then it costs US$3 to US$5 per server a month.
Model families
You can run models from these families on xCloud's plans:
- Llama (Meta): Llama 3.2 1B and 3B, and Llama 3.1 8B
- Gemma (Google): Gemma 3 1B, 4B, 12B and 27B
- Qwen (Alibaba): Qwen3 from 0.6B to 14B, plus Qwen3 30B-A3B
- DeepSeek: DeepSeek-R1 distills at 7B, 8B and 14B
- Mistral: Mistral 7B and Mistral Small 3 (24B)
- gpt-oss (OpenAI): gpt-oss-20b
- Phi (Microsoft): Phi-4
The largest models, such as Gemma 3 27B, Qwen3 30B-A3B and Mistral Small 3, need the 24 GB Ollama plan.
New on the site: the Phi family
Alongside xCloud, we've added a new model family to LLM Hosting: Phi. Phi is a family of small language models developed by Microsoft. The idea behind Phi is that a small model trained on carefully chosen, high-quality data, much of it synthetic, can match models many times its size.
The first Phi model on the site is Phi-4. It has 14 billion parameters and is released under the permissive MIT licence, so you can use it in commercial projects. Phi-4 is especially strong at maths, logic and step-by-step reasoning, and it also does well at coding help and answering questions.
Because it is small, Phi-4 is a good choice when you want solid reasoning on modest hardware. It runs on both of xCloud's CPU plans, and suits tasks like tutoring and maths help, analysing documents, drafting code, or powering a reasoning-heavy assistant without paying for a GPU.
See the listings
You can compare xCloud's plans with every other host on the site, and filter by model, price and RAM to find the right fit for your project.
To get started, visit xCloud's website.
Sources
Related model families
LLM hosts mentioned
| Host | LLM hosting plans | Model families |
|---|---|---|
| xCloud | xCloud Ollama Hosting, xCloud LocalAI Hosting | Llama, Gemma, Qwen, Mistral, Phi, DeepSeek, gpt-oss |
