OpenAI just cut GPT-5.6 Luna’s price by 80%. ChatGPT crossed a billion weekly users. Google quietly renamed Gemini Advanced to “Google AI Ultra” and dropped its top tier from $249.99 to $99.99. Every few weeks, one of these four tools changes its pricing, and every few weeks someone asks us the same question: which one should I actually pay for?
Not “which is smartest” — the honest answer to that changes monthly and doesn’t matter as much as people think. The real question is which $20 a month (or $0) actually fits how you work. Here’s the practical answer, plus every price you need to make the call yourself.
What each one is actually built for
Before the numbers, the positioning matters more than the benchmarks:
- ChatGPT is the generalist. Best voice mode, best image generation built in, the widest range of “just ask it anything” use cases. It’s the one most non-technical people should start with.
- Claude is the one built for depth — long documents, careful reasoning, and coding. If your work involves reading, writing, or reviewing anything long-form, this is where it shows.
- Gemini lives inside Google. If your day already runs through Gmail, Docs, Sheets, and Android, Gemini shows up where you’re already working instead of asking you to open another tab.
- Perplexity is a research tool first, a chatbot second. Every answer comes with sources attached, which the other three treat as an afterthought.
The pricing, side by side
| Tool | Free | Mid tier | Top tier |
|---|---|---|---|
| ChatGPT | $0 (Free) / $8 (Go) | $20/mo (Plus) | $100–$200/mo (Pro) |
| Claude | $0 (Free) | $20/mo (Pro) | $100–$200/mo (Max) |
| Gemini | $0 (AI Studio) / $4.99 (AI Plus) | $19.99/mo (AI Pro) | $99.99–$199.99/mo (AI Ultra) |
| Perplexity | $0 (Free) | $20/mo (Pro) | $200/mo (Max) |
Notice the pattern: all four converge on $20/month for their “serious user” tier, and all four have a $100–$200 tier aimed at people running heavy daily workloads or coding agents, not casual chat. Unless you’re already hitting daily limits on a $20 plan, the top tiers aren’t for you yet.
The best AI tool isn’t the smartest one on a benchmark — it’s the one that’s already open when you need it.
How we actually use them
For what it’s worth — and this is just our own workflow, not a universal rule — Claude is what we reach for when we’re actually researching something, ChatGPT is what we open for story and copy writing, and Google’s AI handles the quick technical lookups, especially anything app-development related. None of that means those are objectively the “best” choices for those jobs; it’s just what stuck after enough daily use. Your own workflow will probably look different, and that’s fine — the point of the table above is to let you test the $20 tier of whichever one matches your actual use case before assuming you need to pay for all four.
Which one should you actually pay for
- Students and researchers: Perplexity Pro or Claude Pro. Perplexity if you need citations for every claim; Claude if you’re synthesizing long readings into your own writing. If you’re on a tighter budget, our guide to the best free AI tools for students covers what you can get done without paying anything.
- Writers and creators: ChatGPT Plus. The voice mode and image generation alone justify it, and it’s the most forgiving for open-ended brainstorming.
- Developers: Claude Pro, moving to Max if you’re running Claude Code daily. It’s consistently the strongest of the four for reading and writing real codebases, not just snippets.
- Anyone deep in Google Workspace: Gemini AI Pro. The $19.99 tier’s 1M-token context window plus native access inside Docs, Sheets, and Gmail beats context-switching to another app, even if the model itself isn’t always the sharpest.
If you only pay for one, start with whichever matches the job you do most often — not the one with the best headline benchmark score this month. You can always add a second $20 subscription later; almost nobody needs three.
Frequently asked questions
It depends on what you use it for, not which one benchmarks highest. ChatGPT Plus suits general writing and brainstorming, Claude Pro suits coding and long-document work, Gemini AI Pro suits anyone living inside Google Workspace, and Perplexity Pro suits research that needs citations. All four cost $20/month at the serious-user tier.
Neither is universally better. ChatGPT Plus has stronger voice mode and built-in image generation, making it more forgiving for general use. Claude Pro is stronger at reading and writing long documents and real codebases, making it the better pick for developers and heavy writers.
Almost nobody needs three. Start with the $20/month tier of whichever tool matches the job you do most often, and only add a second subscription if you consistently hit a real gap — for example, pairing Claude Pro for work with Perplexity Pro for citation-backed research.
ChatGPT Plus at $20/month includes GPT-5.6 access. OpenAI cut GPT-5.6 Luna’s API pricing by 80% in 2026, but that price cut applies to developers building on the API, not to the consumer subscription tiers.
Yes, if research and citations matter for your work — Perplexity is built to show sources for every answer, which the other three tools treat as an afterthought. Perplexity also offers an Education Pro tier at $10/month for verified students, half the price of the standard Pro plan.
