About AI Hub

AI Hub is an independent English-language guide to the practical side of AI: open-weight models, the software that runs them locally, cloud AI services, agents and MCP tools, and the hardware behind it all.

We exist to answer the questions people actually ask: which model fits my 12 GB GPU?, Ollama or LM Studio?, is local AI cheaper than the API for my usage? Our interactive finders — Task Finder, VRAM Calculator, Stack Builder — are built for exactly those questions.

What we cover

• Local LLMs — open-weight language, image and audio models: licenses, VRAM needs, real performance.
• Software — runners and tools: Ollama, LM Studio, llama.cpp, vLLM, ComfyUI and friends.
• Cloud AI — ChatGPT, Claude, Midjourney and similar services: pricing, free tiers, APIs.
• Agents & MCP — agents, plugins, MCP servers and skills, and how they connect to your models.
• Hardware — GPUs, PCs, Macs and workstations for local AI, with VRAM and price data.

How we test

Reviews are based on hands-on testing where possible: we install the software, run the models, and measure what matters — speed, memory usage, quality of output, and how hard it is to set up. Where we cannot test something directly, we say so, and we lean on the vendor's documentation and reproducible community benchmarks.

Ratings are editorial: a 5-point scale that weighs ease of use, performance, cost and documentation. We update pages when versions change — check the Last updated date on each review.

Sources & advertising policy

The site is supported by advertising. Advertisements are clearly marked and never influence reviews, ratings or recommendations. Where a link sends you to a vendor's site, it may be an affiliate or outbound link marked nofollow.

Technical claims (VRAM requirements, benchmark numbers, licensing) are cited to vendor documentation, model cards, or linked sources within the relevant article.