About The AI Standings
Hi, I'm Alex. I've worked in enterprise IT for more than 20 years, and these days a lot of that work is about AI. This site is where I share what I learn with everyone else.
Who's behind this
I've spent more than two decades in enterprise IT. Today I lead an IT support team at a large healthcare organization, where my job includes deciding which tools are safe and worth adopting, and training clinical departments to use AI responsibly. Healthcare is a demanding place to test technology: privacy rules are strict, mistakes are expensive, and the people using the tools don't have time for hype.
Outside my day job I run ThinkXELA, an independent IT consultancy that gives small businesses honest, vendor-neutral technology advice. A growing share of the questions I get there are some version of "which AI should we use, and what will it cost?"
The AI Standings is my answer to that question, updated every week. I follow model launches, read the benchmark reports and pricing pages, and use the models myself for real work: writing, analysis, coding and building small tools. Then I write it up in a format that's quick to read and honest about what each model does well and badly.
How I use these models
I don't just read about AI models. They're part of my working week:
- At work, I build and test prompts that help teams analyze operational data, and I teach staff what they should and shouldn't paste into a chatbot.
- For clients, I use AI assistants to plan and build websites and small business apps, which is where the differences between coding models show up fast.
- For this site, I weigh what I see in that day-to-day use against the public benchmarks and leaderboards before a model moves up or down.
When a guide says a model is good or bad at something, it's based on that use plus the public evidence, and I link the evidence so you can check it.
Why standings?
In 2026 a new "best model in the world" ships almost every week. In one Tuesday in September, Anthropic, OpenAI and Xiaomi released five models between them. Nobody has time to read five launch posts and a dozen benchmark tables.
So I do it the way sports pages cover a season. Every week there's a new edition of the standings. It shows who's on top in each category, who moved up, who fell, and what it means for you.
What's here
- Weekly power rankings for general AI assistants, coding, value for money, open-weight models, image generation and video generation, with an archive of past editions.
- Model profiles with prices, strengths, weak spots and ranking history.
- Plain-English guides on using AI safely and well for work, side income, creativity and everyday life.
- Free tools: a Which AI should I use? quiz, a head-to-head comparer and an API cost calculator.
How the rankings are made
Rankings combine public benchmark results, published pricing, independent leaderboards and hands-on use. The full approach, including how much each factor counts, is on the how we rank page. Rankings are editorial opinion, and I show my reasoning so you can disagree with it.
Independence
The site is supported by advertising. No AI company pays to be ranked, none can pay to move up, and I don't take affiliate commissions for recommending a model. If sponsored content ever appears, it will be clearly labelled.
Corrections and feedback
AI moves fast, and mistakes happen. If you spot an error or think a ranking is badly wrong, please get in touch. Errors get fixed quickly and corrections are noted.
Last updated October 26, 2026