What is Value×? How to compare AI plans by real compute per dollar
Most AI plan comparisons stop at the price. But two plans at the same price can differ wildly in how much you actually get. Value× is the metric we use to cut through that — and this post explains exactly what it means.
The one-line definition
Value× is an estimate of the API-equivalent model usage a subscription unlocks, divided by its monthly price.
A plan with a higher Value× gives you more usable compute per dollar. A 10× plan unlocks roughly ten dollars of API-equivalent usage for every one dollar you pay.
Why not just compare prices?
Because price tells you what you spend, not what you get. Consider two $20 plans:
- Plan A includes a small number of premium requests, then hard-stops.
- Plan B includes a generous usage pool covering frontier models.
They cost the same. But Plan B might let you do 3–5× more real work before hitting a limit. Comparing on price alone would call them equal. Value× makes the difference visible.
How it's estimated
For each plan we look at what the subscription includes — the models, the included usage (requests, tokens, or a dollar-denominated pool), and the limits — and estimate what it would cost to buy the equivalent usage at standard API rates. Then:
Value× = (estimated API-equivalent value per month) ÷ (monthly price)
A few important notes about the estimate:
- It's an estimate, not a guarantee. Real-world value depends on how you use the plan. We label how confident each figure is (measured, inferred, or self-reported).
- Free plans don't get a multiplier — dividing by a $0 price is undefined. We rank those qualitatively instead.
- We confirm prices and limits against provider sources and date-stamp them, because this market moves monthly.
How the gauge reads
On the comparison tool, Value× is shown on a log scale against a reference point, so a 2× plan looks meaningfully different from a 17× plan. A linear scale would crush everything below the top performer into a flat sliver and hide the differences that matter.
How to use it
- Set your budget — pick a tier (e.g. under $20, under $50, under $100).
- Sort by Value× — the highest-leverage plan in that budget floats to the top.
- Sanity-check the fit — the best Value× plan still has to suit your workflow (editor vs agent vs raw API). Open its provider page to read the details.
The takeaway
Price is the question everyone asks first, but it's the wrong place to stop. The plan that's best value is the one that unlocks the most usable compute for what you pay — and that's what Value× is built to surface.