The list ladder for the Oracle Java SE Universal Subscription runs from 15.00 dollars per employee per month at the smallest sizes to 5.25 dollars per employee per month at the very largest. On paper that looks like a volume discount. In practice it is the part of the model buyers most often misread, because the rate that falls is multiplied by a number that keeps growing: your entire counted workforce.
This article decodes the ladder rung by rung, shows what each rate implies in real dollars, and explains where a buyer can move the number. The short version is that the rate is negotiable, the band you land in is negotiable, and the population the rate multiplies is the single biggest lever you control.
Why a falling rate still raises the bill
The metric counts every full time and part time employee, every contractor, and every temporary worker, regardless of whether any of them ever touches Java. So the bill is the counted population multiplied by the per employee rate multiplied by twelve months. As you grow, the rate steps down through bands, but the population grows faster than the rate falls. A company that doubles its workforce does not halve its rate. It moves down perhaps one band while the headcount it pays for doubles. The net effect is a larger annual number on a lower unit price.
This is the central illusion of the ladder. Oracle presents the lower rate as the reward for scale. The buyer who only looks at the per employee figure feels reassured. The buyer who multiplies it out sees the real trajectory.
The 2026 ladder, indicative
The table below shows indicative list rates by band and the annual list cost at the top of each range. Negotiated rates differ, and the band boundary itself can be negotiated. Treat every figure here as indicative.
| Band, employees | Indicative list rate per employee per month | Indicative annual list at band top |
|---|---|---|
| 1 to 999 | $15.00 | $179,820 |
| 1,000 to 2,999 | $12.00 | $431,856 |
| 3,000 to 9,999 | $10.50 | $1,259,874 |
| 10,000 to 19,999 | $8.25 | $1,979,901 |
| 20,000 to 39,999 | $6.75 | $3,239,919 |
| 40,000 and above | $5.25 | scales upward |
Read the table by the shape, not the cells. The unit price drops by almost two thirds from the top rung to the bottom, while the annual cost climbs into the millions. The rate is the headline. The population is the engine.
What each rung means for a buyer
At the 15.00 rung, small estates pay the ceiling rate on a small population. The pain is the unit price, and the lever is whether you need a commercial Oracle Java subscription at all or whether a free OpenJDK distribution covers your workloads. At the middle rungs, the population is large enough that the total is a serious budget line, and the lever shifts to bounding who is counted. At the 5.25 rung, the unit price is as low as the model goes, so the only meaningful lever left is the size of the counted population and the contract terms around it.
The practical takeaway is that the most expensive rung to negotiate is whichever one you have not yet challenged. The number to attack is rarely the rate alone.
The boundary trap between rungs
Bands create a quiet hazard at their edges. A company sitting just above a boundary pays the lower rate across its whole population, while a company just below pays the higher rate on a smaller one. That sounds fair until an annual true up pushes you across a boundary as you hire. You can move up a band, lifting the rate base, while the count rises at the same time. The two effects compound. A buyer watches the boundary as closely as the rate, because a modest headcount swing near an edge can move the effective price more than a hard fought negotiation on the rate itself.
Indicative worked example. A services firm with 9,800 counted employees sits at the top of the 3,000 to 9,999 band at 10.50 dollars. Hiring 300 people pushes it into the next band. The rate falls to 8.25, but the count it pays for is now 10,100. The annual list moves from roughly 1.23 million to roughly 1.00 million only because the lower rate offsets the higher count. Near a boundary, small moves produce counterintuitive results. Model them before you sign.
The levers that actually move the number
Three levers matter. The first is the counted population. Most estates carry headcount that has no business being inside a Java envelope, and bounding that population with evidence is the largest single saving available. The second is the deployment itself. By isolating Oracle Java to the workloads that truly require it and migrating the rest to a free OpenJDK distribution, you shrink the envelope so the band you negotiate is the smallest defensible one. The third is the contract. Fixing the term, capping the counted population, and stripping the true up stop the ladder from quietly walking you upward year after year.
For the full picture of how the bands are structured and where the edges sit, read our companion piece on the 2026 Universal Subscription pricing bands. To understand the offer the rate sits inside, read how the Oracle Java SE Universal Subscription actually works.
The buyer side next step
The ladder is an input you can shape, not a fixed fact handed down. Before you accept any rung, model your real exposure from evidence and decide how much of your estate genuinely needs an Oracle Java subscription at all. Our Oracle Java Licensing Guide for 2026 sets out the full landscape, and the buyer side defense always starts with a number you can trust. We work on a Fixed Fee from $18,000 or a Gainshare share of verified savings or avoided exposure, with zero retainer and no risk to you. Across the estates we defend, the average reduction is 68 percent versus Oracle's opening number, and we confirm your figure before you commit.
Next step. Download the Oracle Java Audit Survival Guide for the complete buyer side playbook, or get a quote below and we will rebuild your exposure from evidence.