Guides · · 4 min read
Plainly vs Perplexity: two different tools
Perplexity is an answer engine: search with citations. Plainly is an ask-anything assistant for $5 a month. Where each wins.
Articles on getting more from cheap AI models
Plain-English guides to DeepSeek, open models, and cheap ChatGPT alternatives — written by Tom Hill, Works at an AI startup.
Interactive · 5 minutes
The Blind Test
Everyday writing tasks went to ChatGPT's Go plan, Plainly, and a frontier model. Read the answers — can you spot the frontier?
Guides · · 4 min read
Claude Pro usage limits: what runs out
Claude Pro costs about $20 a month and its limits arrive in practice rather than on paper: heavy use runs out, and Max starts at $100. Who should not pay.
Guides · · 4 min read
How do AI apps make money? The $5 answer
AI apps earn from subscriptions, ads, training data, API markups or credit metering. Plainly earns with the subscription alone: the arithmetic behind $5.
Guides · · 3 min read
Why is there no free AI plan? Someone pays
Free AI plans are not sustainable: free users are funded by ads, by data used to train models, or by other people's subscriptions.
Guides · · 5 min read
How to switch from ChatGPT: what you keep
Switching from ChatGPT takes about ten minutes. Export what you need, expect to rebuild your context, and keep the old account for a while.
Guides · · 7 min read
The best AI apps, ranked on what they cost
Ranked on one criterion: what you pay against the work you actually do. Six apps, honest verdicts, the two I would not recommend.
Guides · · 4 min read
What's the catch with cheap AI? A scope limit
Cheap AI is real, and so is the catch. You are on value models rather than frontier intelligence: fast answers to questions, and not much beyond them.
Guides · · 6 min read
What does LLM mean? An AI glossary to ignore
LLM means large language model, not the law degree. Every AI term an ordinary person trips over, in one plain line each, and why none of them matter.
Guides · · 3 min read
Is Plainly right for me? Who it fits
Plainly fits if you want one clean window, separate chats and $5 a month. Not if you need API access, model choice, images or cross-chat memory.
Guides · · 5 min read
ChatGPT for students: the four free months
Eligible US students can claim four free monthly billing periods of ChatGPT Plus, worth $80, before 31 October 2026. It renews at $20 unless you cancel.
Guides · · 4 min read
Plainly vs ChatGPT: where each one wins
Plainly is $5 a month for one clean AI window. ChatGPT is faster and smoother, and its free tier is a good deal. Price, features, and who stays put.
Guides · · 6 min read
ChatGPT Plus vs Pro: what $100 to $500 buys
Pro runs $100 to $500 against Plus's $20, and the difference that matters is memory: 128,000 tokens against 54,000. Every Pro tier is ad-free.
Guides · · 5 min read
Is ChatGPT Plus worth it? Only for the frontier
ChatGPT Plus is worth $20 for people who use the frontier most days for hard work. It does not lengthen the conversation, as Plus shares Go's window.
Guides · · 8 min read
Is ChatGPT Go worth it? What $8 buys over free
ChatGPT Go is worth $8 for a doubled context window and higher tool limits. It does not buy the better model, remove the ads, or fix the forgetting.
Guides · · 5 min read
Best cheap ChatGPT alternatives, ranked
Ranked on one criterion: a flat monthly price for full-capability answers, no credit packs, no ads, and no data taken as payment.
Guides · · 4 min read
What is a cheap ChatGPT alternative?
A flat-price AI app that serves a full-capability model for around $5 a month. What counts, what does not, and where cheap stops.
Guides · · 2 min read
Does Plainly train on your chats?
No. Plainly does not train models on your chats and no human reads them. Every question routes through a zero-data-retention provider.
Guides · · 8 min read
What's the best AI for writing? A blind test
Three models, five writing tasks, blind scoring, first response only. A $5 app won the email. ChatGPT served ads against three of its five answers.
Guides · · 3 min read
DeepSeek app vs ChatGPT app: the honest comparison
Both work. ChatGPT is the smoother product with ads below $20; the DeepSeek app is capable but locks you to one lab and retains your prompts.
Guides · · 2 min read
The future of AI pricing: cheaper, not pricier
Prediction, September 2026: China keeps shipping hyper-efficient models and takes the low-to-medium market; the US stays at the frontier.
Guides · · 3 min read
GPT-OSS vs DeepSeek: two different stories
Both carry the open-weights label; only one is useful infrastructure. GPT-OSS is okay and I believe it was a legal play. DeepSeek-class weights are real.
Guides · · 3 min read
Have Chinese AI models caught up? The 2026 picture
At the low and middle end they own the best cost-to-output in the market, and a new release lands almost weekly. The US frontier still wins the very top.
Guides · · 2 min read
Why are there so many AI apps?
A new capable model lands almost weekly and every lab wraps its own app around it. Keeping up is a full-time job, which is why you should not try.
Guides · · 3 min read
Free vs $5 vs $20 AI apps: what each gets you
Free tiers of ChatGPT and Claude are demos: capped, small-context, ad-supported. A $5 subscription covers everyday use; $20 buys frontier access few need.
Guides · · 5 min read
What AI should I recommend to a friend?
Recommend a cheap, simple app that picks the model for them. One login, no setup, no choices. The exact script for when a friend asks what AI to use.
Guides · · 4 min read
Is a $5 AI subscription good enough? Yes
For everyday use, yes: unlimited chats on capable models. Not if you need frontier Western models for intense work. Both parts stated plainly.
Guides · · 4 min read
Why China open-sources AI and the US does not
China pushes its labs to release open weights; the US keeps its frontier closed because access is the business. That difference explains AI pricing.
Guides · · 4 min read
Why are Chinese AI models cheaper? Necessity
Chinese AI models are cheaper out of necessity: less funding, chip export bans and brutal domestic competition forced the labs to get efficient fast.
Guides · · 4 min read
Is there one best AI model? No, and here's why
There is no single best AI model, so power users dual-wield: frontier models for high-stakes work, cheap open-source models for everything else.
Guides · · 4 min read
Why is ChatGPT so expensive? The cost structure
ChatGPT was never built to be cheap. Billions in funding, the most advanced chips available, no cost pressure. The $20 price is a side effect.
Guides · · 4 min read
Which AI model should I use? None of them
You should not have to pick an AI model. The frontier changes every few days and each model is good at something different. Choosing is the failure mode.