Practical guide · verified against the real thing
ChatGPT vs Claude vs Gemini: an honest, non-hyped comparison
In one line: Pricing verified, model lines named without version-number theatre, and the strengths stated as trade-offs — including the ones each company would rather not headline.
Every AI comparison is a time capsule, so this one opens with its expiry date: written September 2026, prices verified against the providers' own pages and the model lines described by name — not version number, because the versions rotate quarterly while the page you're reading doesn't. What survives the rotation is each assistant's shape: what it's good at, what it costs, and what its free tier is for.
The pricing, side by side (verified September 2026)
ChatGPT (OpenAI): Free tier; Go at $8/month; Plus at $20; Pro at $100 and a $200 tier. Claude (Anthropic): Free tier; Pro at $20 (about $17 monthly on annual billing); Max at $100 and $200 for the heavy usage tiers. Gemini (Google): free tier; AI Plus at $4.99 — the cheapest paid entry of the three; AI Pro at $19.99; AI Ultra at $99.99 and $199.99. All three bundle extras differently — Google's tiers ride on One storage, OpenAI's upper tiers carry the creative tools and highest limits, Anthropic's carry expanded usage of its coding and agent tooling. The one honest buying note: free tiers first. A week of real use costs nothing and beats every review (including ours — here's where the free tiers end).
The shapes, honestly
ChatGPT is the everything-store: voice, image generation and analysis, the widest third-party plugin ecosystem, and the best-polished mobile apps. Its weakness is the flip side of being everything — the assistant that's part of a hundred features is rarely the deepest at any one, and the free tier's flagship access is the most metered of the three. Claude is the writer's and builder's pick: long documents handled natively, prose that needs less editing, and — for the technical — the coding and agent tooling that professionals have quietly adopted fastest. Its free tier runs on rolling usage windows that run out sooner than you'd like on a working day. Gemini is the context-length and integration play: enormous document windows, and living inside Gmail, Docs and Search where your stuff already is — with the cheapest paid entry price. Its weakness is the same integration: your use of it is woven into Google's account ecosystem, and the data settings deserve a read (we checked the training-settings question here).
How to choose without the hype
Ignore benchmarks for a week; they measure a moving target and your prompts aren't the benchmark's prompts. Instead: take one real task you do weekly — a document you draft, code you maintain, research you sift — and run it through all three free tiers. The winner for your work is usually obvious inside five prompts, and the paid question only matters after the free tier has earned it. Our deeper dives: Gemini vs ChatGPT, the DeepSeek value case, what Arena's head-to-head testing actually shows, and the wider alternatives map.
Pricing verified September 2026 against OpenAI, Anthropic and Google's own plan pages; model lines described by name (versions rotate quarterly — check the provider's model page on the day). This page gets a refresh check before every republish.
Sources
Next