I went looking for a straight answer to this and came back with a folder full of articles that were all written for somebody else. Most of them were written for a person with a salary, an employer, and a payroll department. A few were written for people with several hundred thousand dollars of profit and an accountant on retainer. Almost none were written for one person with a Schedule C and a question.
So this is the version I wanted to read. Two accounts, what separates them, and the part that turns out to matter more than the numbers everyone leads with.
The part I said was a different conversation
In the post about the tax-savings account I never touch, I drew a hard line around what that account is for. It holds money that leaves four times a year. It is short-term storage for money that was never mine. I said that where your long-term savings go was a different decision with its own tradeoffs, and a different conversation.
This is that conversation.
The line still matters, because the two things get confused constantly. The tax account is a holding pen. A retirement account is the opposite: money you are deliberately locking away from yourself, with a penalty attached to changing your mind. One is plumbing. The other is a decision. If you have not built the plumbing yet, build that first. There is no version of this post that helps someone who is still surprised by their quarterly bill.
Assume that part works. Assume there is money left after the set-aside, and you have started wondering where it should go instead of sitting in checking. The two answers most freelancers land on are a Solo 401(k) and a SEP-IRA.
The ceilings, since that's why you're here
You came for the limits, so here they are for 2026.
Solo 401(k). You wear two hats. As the employee, you can defer up to $24,500. As the employer, you can add a profit-sharing contribution on top. Everything together stops at $72,000. If you are 50 to 59 or 64 and older, a catch-up of $8,000 raises your stop to $80,000. If you are 60 to 63, an enhanced catch-up of $11,250 raises it to $83,250.
SEP-IRA. One hat. You contribute up to 25% of compensation, stopping at $72,000. There is no catch-up at any age. Turning 55 does not change your number.
Both of those went up for 2026. The shared cap was $70,000 last year, and the deferral was $23,500. If you read something on this topic written before last November, its numbers are wrong.
Now look at what that comparison actually gave you. Both plans stop at $72,000. The number everyone puts in the headline is the number the two plans have in common. It is the least useful fact in the whole discussion, and it is the reason these comparisons tend to end with "it depends."
One aside for anyone paying themselves a W-2 salary through an S-corp. Starting in 2026, if you had more than $150,000 in FICA wages from your own company last year, your catch-up contributions have to go in as Roth. You lose the deduction on that slice. If you are a sole proprietor with no payroll, this does not reach you, because you pay self-employment tax on net earnings rather than FICA on wages.
What you can actually put in
Here is the thing that took me longest to find stated plainly.
That "25% of compensation" for a SEP-IRA is not 25% of your Schedule C net profit. Compensation, for this purpose, means your net earnings from self-employment, which is your net profit reduced by the deductible half of your self-employment tax. And because the contribution itself comes out of the same figure, the arithmetic collapses from 25% into roughly 20% of net earnings, which lands near 18.6% of net profit.
Nearly every article about SEP-IRAs says 25%. For anyone reading this, 25% is the wrong number. It is the number for an employee whose employer runs the plan. If you file a Schedule C, you are not that person.
Run it at a round number. Say you cleared $100,000 of net profit.
SEP-IRA: about $18,600. Not $25,000.
Solo 401(k): your $24,500 employee deferral, plus an employer contribution figured the same way as the SEP, so about $18,600 on top. Call it $43,100.
Same income. Same $72,000 headline ceiling on both. Two and a half times the room in one of them.
That gap is the actual answer for most people, and it exists because the Solo 401(k) lets you contribute as an employee and as an employer, while the SEP only ever sees you as the employer. The shared $72,000 cap only starts to bind somewhere in the high six figures of profit, which is to say it is irrelevant to almost everyone who searches for this comparison.
Two notes before anyone runs their own numbers. The self-employment tax piece shifts a little once your earnings pass the Social Security wage base, so treat 18.6% as close rather than exact. And all of it starts from net profit, which means it starts from having categorized your year correctly. What counts as what on Schedule C decides the input to every number above, and getting it wrong quietly moves your ceiling. Worth saying that the reason I know my net profit at any point in the year is that I keep a running total, not that I am good at arithmetic in April.
Also worth saying: what you can afford to put away is downstream of what you charge. No account structure fixes a rate that leaves nothing at the end of the year.
The deadline advice you'll read is out of date
This is the part that surprised me, and it is the reason the comparison usually gets framed wrong.
The standard advice is that a SEP-IRA is the flexible one and a Solo 401(k) has to be open by December 31. That was true. It stopped being true in 2023, and a lot of what is still online was written before the change.
The SEP-IRA is genuinely flexible. You can open one and fund it as late as your tax filing deadline, extensions included. You can decide in September of the following year, file an extension, and still make a contribution for the year that already ended. For a business with lumpy income, that is a real advantage and it has not changed.
The Solo 401(k) is less rigid than it looks. Section 317 of SECURE 2.0 lets a sole proprietor with no employees adopt a plan and make first-year employee deferrals up to the tax filing due date, without extensions. For the 2026 tax year, that means roughly April 15, 2027, not December 31, 2026. The December 31 wall that half the internet warns about is, for a first-year sole proprietor, no longer where it was.
Two qualifications, because this is where it gets easy to hurt yourself. The extended first-year window covers the first plan year only. Once the plan exists, deferrals for later years have to be elected by December 31 of that year, so the December 31 deadline is a returning-year constraint rather than a starting one. And providers differ on whether they will actually process retroactive first-year deferrals, so the rule permitting it does not guarantee your provider supports it. Ask before you assume.
The employer side of both plans follows the filing deadline with extensions, so that piece is a wash.
How I'd think about it
I am not going to tell you which one to open, because the honest answer depends on facts about you that I do not have. But the decision is narrower than it looks.
If you want the most room at a normal freelance income, the Solo 401(k) is the one with more room, by a lot, and the gap widens the more you earn up to the point where the shared cap binds. If you want the least paperwork and the longest runway, the SEP-IRA opens in an afternoon, has no plan document to maintain, and waits for you until October. If you are 50 or older, only one of them has a catch-up.
If you want a default rather than a framework: for one person with no employees and no plans to hire, the Solo 401(k) is the one I would look at first, on the strength of that contribution gap alone. The SEP-IRA earns its place when simplicity or a late decision matters more to you than the ceiling does, and for a lot of people in a bad year, it will.
And if you ever hire someone, the calculus changes underneath you, because a SEP requires you to contribute the same percentage for eligible employees that you contribute for yourself. That is a good problem to have and a bad one to discover in March.
The thing I would not do is spend another evening comparing the $72,000 figures, which is what most of the articles invite you to do. They are the same number. Start from your net profit, look at what each plan would actually let you contribute at that number, and then look at how much time you have left to decide.
This is also a reasonable place to spend an hour of a CPA's time rather than a weekend of your own. Bring your net profit, your age, and whether you intend to hire anybody. That is most of what they need to answer it, and it is a much shorter conversation than the one you will have if you show up with a shoebox.
