Starter 401(k) vs. SIMPLE IRA for Freelancers (2026): The New Plan With Zero Mandatory Employer Cost and a Much Lower Cap

Published: October 2, 2026 · Reading time: 11 min

TL;DR: SECURE 2.0 created a genuinely new retirement plan — the Starter 401(k) deferral-only arrangement under IRC §401(k)(16) — for any employer with no other retirement plan. It can never require an employer contribution, by statute, but the trade is a deferral cap on its own, slow-moving schedule: just $6,000 for 2026 — unchanged since 2024 and now actually below the regular IRA limit of $7,500 — plus a $1,100 catch-up at age 50+ that does mirror the IRA catch-up figure exactly, both confirmed directly against IRS Notice 2025-67. A SIMPLE IRA under IRC §408(p) lets employees defer far more — $17,000 for 2026 ($18,100 at eligible small employers) — but requires a mandatory 3% match or 2% nonelective contribution every year, with no discretionary skip. In the worked example below, a freelance designer netting $90,000 — who qualifies for the SIMPLE's enhanced $18,100 small-employer limit because she's never maintained a retirement plan of her own (a freelancer winding down her own prior Solo 401(k) would not qualify, and would be stuck with the standard limit instead) — shelters $20,593.45 of her own money through a SIMPLE versus $6,000 through a Starter 401(k) — worth $1,400.97 more in current-year tax savings once the §199A taxable-income cap claws back part of it — against a $560–$840 mandatory employer cost for her new hire that the Starter 401(k) avoids completely. The two plans are legally mutually exclusive, and a solo freelancer with no employees gets nothing from a Starter 401(k) that a plain personal IRA doesn't already give them.

Every freelancer retirement-plan guide on this site so far has compared plans that have existed for years — SEP, SIMPLE, Solo 401(k). This one is genuinely new. SECURE 2.0 added an entirely different plan type in late 2022, effective for plan years beginning after December 31, 2023, and it inverts the usual trade-off: instead of a higher cap bundled with more employer obligation, it offers a vastly lower cap bundled with a statutory guarantee that the employer will never owe anything at all.

The Statute, Read Directly

IRC §401(k)(16) is titled "Starter 401(k) deferral-only plans for employers with no retirement plan." Fetched directly from the Code:

"(D) Contribution limitations (i) In general The requirements of this subparagraph are met if, under the arrangement— (I) the only contributions which may be made are elective contributions of employees described in subparagraph (C), and (II) the aggregate amount of such elective contributions which may be made with respect to any employee for any calendar year shall not exceed $6,000."

Two things in that text matter more than they look. First, clause (I) isn't a cap on the employer's contribution — it's a flat prohibition on the plan having one at all. There's no "unless the employer elects to match" escape hatch anywhere in the subsection. Second, the dollar figure in clause (II) isn't indexed the way a regular 401(k)'s $24,500 2026 deferral limit is (that one tracks §402(g)(4)'s own schedule). The very next clause sets a different, lower indexing path:

"(ii) Cost-of-living adjustment In the case of any calendar year beginning after December 31, 2024, the $6,000 amount under clause (i) shall be adjusted in the same manner as under section 402(g)(4), except that '2023' shall be substituted for '2005'."

That's a deliberately slower-moving number than the regular 401(k) limit — and IRS Notice 2025-67, the government's own 2026 cost-of-living adjustment notice, confirms it hasn't moved at all yet:

"The limitation under section 401(k)(16)(D)(i)(II) and 403(b)(16)(D)(i)(II) that generally applies for elective contributions made to a starter 401(k) deferral-only arrangement described in section 401(k)(16)(B)... remains $6,000. This limitation is increased for individuals who attain age 50 before the end of the taxable year by $1,100."

So for 2026: $6,000 flat, $7,100 at 50 or older. The $1,100 catch-up isn't a new Starter-401(k)-specific number — it's defined by cross-reference to §219(b)(5)(B)(ii)'s IRA catch-up amount, the same $1,100 figure the IRS just confirmed for ordinary IRA contributions this year. There's no separate, larger catch-up for ages 60–63 the way a regular 401(k) or a SIMPLE provides; the Starter 401(k)'s catch-up structure mirrors an IRA's exactly, not a 401(k)'s.

What a SIMPLE IRA Requires That a Starter 401(k) Categorically Can't

A SIMPLE IRA's matching mechanic sits in IRC §408(p)(2)(A), read directly:

"(iii) the employer is required to make a matching contribution to the simple retirement account for any year in an amount equal to so much of the amount the employee elects under clause (i)(I) as does not exceed the applicable percentage of compensation for the year."

"Required" is the operative word — not "may elect to." §408(p)(2)(C)(ii) defines the applicable percentage as 3% by default, with a narrow ability to drop it to as low as 1% in up to 2 of the preceding 5 years, never lower, never for longer. The one real alternative is the 2% nonelective design under §408(p)(2)(B)(i):

"An employer shall be treated as meeting the requirements of subparagraph (A)(iii) for any year if, in lieu of the contributions described in such clause, the employer elects to make nonelective contributions of 2 percent of compensation for each employee who is eligible to participate in the arrangement and who has at least $5,000 of compensation from the employer for the year."

The 2% nonelective version is actually the stricter cost in one sense: it's owed to every eligible employee whether or not they personally defer a dime, where the 3% match is only owed on top of what an employee actually chooses to contribute. Either way, the match or nonelective contribution is written into §408(p)(2)(A)(iii)'s own definition of a "qualified salary reduction arrangement" — it isn't an optional sweetener layered on top; an arrangement that skips it in a given year simply isn't a qualified SIMPLE for that year.

The 2026 Numbers, Verified

Every figure above is read directly from IRS Notice 2025-67 (the government's own 2026 cost-of-living-adjustment notice) and the underlying Code sections, fetched for this article rather than carried forward from an earlier year's guide.

The Enhanced $18,100 Limit Isn't Automatic

It's easy to read "small-employer limit" and assume any small employer gets it. The statute attaches a second, separate condition that trips up exactly the reader most likely to be reading this guide: a freelancer who's run her own Solo 401(k) and is now transitioning away from it to hire a first employee.

§408(p)(2)(E)(i)(I) ties the $18,100 figure to an "eligible employer described in clause (iii)" — the ≤25-employee headcount test. But a separate clause, §408(p)(2)(E)(iv), adds a second, independent condition before an employer counts as eligible for the enhanced amount at all:

"An eligible employer is described in this clause only if, during the 3-taxable-year period immediately preceding the 1st year the employer maintains the qualified salary reduction arrangement under this paragraph, neither the employer nor any member of any controlled group including the employer (or any predecessor of either) established or maintained any plan described in clause (i), (ii), or (iv) of section 219(g)(5)(A) with respect to which contributions were made, or benefits were accrued, for substantially the same employees as are eligible to participate in such qualified salary reduction arrangement."

Clause (i) of §219(g)(5)(A) — the first item on that disqualifying list — is "a plan described in section 401(a) which includes a trust exempt from tax under section 501(a)." A Solo 401(k) is a §401(a) plan. So a freelancer who maintained her own Solo 401(k) at any point in the 3 taxable years before she adopts a SIMPLE does not get the enhanced $18,100 limit for that SIMPLE, even though she now has no more than 25 employees — she's limited to the standard SIMPLE deferral cap ($17,000 for 2026, or whatever the standard figure is the year she actually adopts the plan) until a full 3 taxable years have passed since the Solo 401(k) closed. (The disqualifying list in clause (iv) is narrower than the "qualified plan" list used elsewhere in this guide — it only picks up §401(a) plans, §403(a) annuity plans, and §403(b) annuity contracts, not a prior SEP or SIMPLE — but a Solo 401(k) falls squarely inside it.)

This is exactly why the worked example below is explicit that Riley has never maintained a retirement plan of her own: it's what lets her 3-year lookback come up clean and the $18,100 figure apply without qualification. A reader coming to this guide from Solo 401(k) vs. SIMPLE IRA for freelancers hiring their first employee — i.e., exactly the freelancer winding down an existing Solo 401(k) to bring on a first hire — should run the standard $17,000 limit through the comparison instead, not $18,100.

They Can't Run Side by Side

This is the mechanical point nearly every explainer of the new plan skips. A Starter 401(k)'s own "eligible employer" definition excludes anyone already running one of several other plan types:

"(E) Eligible employer... The term 'eligible employer' means any employer if the employer does not maintain a qualified plan with respect to which contributions are made, or benefits are accrued, for service in the year for which the determination is being made... (iii) Qualified plan The term 'qualified plan' means a plan, contract, pension, account, or trust described in subparagraph (A) or (B) of paragraph (5) of section 219(g)."

Follow that cross-reference to §219(g)(5)(A), and the list of disqualifying plans explicitly includes "a plan described in section 401(a) which includes a trust exempt from tax under section 501(a)" (clause (i)), "a simplified employee pension (within the meaning of section 408(k))" (clause (v)), and "any simple retirement account (within the meaning of section 408(p))" (clause (vi)). A SIMPLE IRA is named by statute, by number, as a plan that locks you out of a Starter 401(k).

The reverse is equally true. A SIMPLE IRA's own exclusive-plan rule, at §408(p)(2)(D)(ii), defines its disqualifying "qualified plan" by cross-reference to the identical §219(g)(5)(A)-(B) list — and a Starter 401(k) is itself "a plan described in section 401(a)," the same clause (i) that names ordinary 401(k)s and Solo 401(k)s. So adopting either one locks out the other, and both lock out an existing SEP-IRA or Solo 401(k) too. This isn't a soft best-practice recommendation — it's the same statutory list cited from both directions.

Eligibility Timing Cuts the Other Way

The dollar-cap comparison favors the SIMPLE IRA heavily. Who gets let into the plan, and how fast, actually favors the Starter 401(k) in the one scenario that matters most for a genuinely new hire — and the compensation test behind the SIMPLE's $5,000 threshold is narrower than it looks:

A freelancer converting an existing 1099 contractor to W-2 status, which is common once a workload justifies the switch, might expect the SIMPLE's lower bar to reach that person quickly. It doesn't: none of the contractor's prior 1099 pay counts toward the $5,000/2-year test, so the clock starts at zero on the W-2 conversion date, exactly as if the person had never worked for the business before. The SIMPLE's faster path only exists for someone who already has 2 qualifying years of actual W-2 compensation on record — a rehired employee, not a converted contractor.

Worked Example: Riley Hires Her First Employee

Riley is a single freelance graphic designer, age 42, Schedule C sole proprietor with no other income and no existing retirement plan — and, importantly for the enhanced SIMPLE limit below, she has never maintained a Solo 401(k), SEP-IRA, or any other retirement plan in the 3 taxable years before this one either. Her 2026 net profit, after deducting Sam's wages and whichever plan cost applies, is $90,000. Sam is a brand-new hire — no prior 1099 or W-2 history with Riley at all — starting at a $28,000/year part-time W-2 salary. That matters for eligibility timing: SIMPLE "compensation" for the $5,000/2-year test counts W-2 wages only (§408(p)(6)(A)(i)'s cross-reference to §6051(a)), so even if Sam had prior 1099 income from Riley it wouldn't count toward it, and as a genuinely new hire he isn't SIMPLE-eligible until he's logged two full years earning $5,000+ — longer than the one year of service a Starter 401(k) requires under its standard §410(a)(1) rule. The mandatory per-employee cost below ($560–$840) is what Riley will owe once Sam clears that bar, not from his first paycheck.

node -e "
const NET_PROFIT_2026 = 90000;
const WAGE_BASE_2026 = 184500; // 2026 SS wage base, SSA determination
const STD_DEDUCTION_SINGLE_2026 = 16100; // Rev. Proc. 2025-32, Sec. 3.14

function seTax(netProfit) {
  const netSE = netProfit * 0.9235;
  const tax = netSE > WAGE_BASE_2026
    ? (WAGE_BASE_2026 * 0.124 + netSE * 0.029)
    : (netSE * 0.153);
  return { netSE, tax, half: tax / 2 };
}

const se = seTax(NET_PROFIT_2026);
const adjustedBase = NET_PROFIT_2026 - se.half;
console.log('net SE earnings (92.35% of profit):', se.netSE.toFixed(2));
console.log('2026 SE tax:', se.tax.toFixed(2));
console.log('deductible half:', se.half.toFixed(2));
console.log('adjusted earnings base (income-tax AGI concept, NOT the SIMPLE match base):', adjustedBase.toFixed(2));
console.log();

// --- Option A: Starter 401(k) ---
const STARTER_2026 = 6000; // IRC Sec.401(k)(16)(D)(i)(II), per Notice 2025-67
console.log('--- Option A: Starter 401(k) ---');
console.log('Riley own deferral cap:', STARTER_2026.toFixed(2));
console.log('Mandatory employer cost for Sam: 0.00 (statutorily prohibited)');
console.log();

// --- Option B: SIMPLE IRA ---
// Riley qualifies for the enhanced small-employer deferral limit under Sec.408(p)(2)(E)(i)(I),
// which cross-references the Sec.408(p)(2)(E)(iii) headcount test: she has no more than 25
// employees who received \$5,000+ compensation in the preceding year (zero, since Sam is her
// first hire -- 'not more than 25' is trivially satisfied). Separately, Sec.408(p)(2)(E)(iv)
// requires that she has maintained NO plan described in clause (i), (ii), or (iv) of
// Sec.219(g)(5)(A) -- i.e. no 401(a)-type plan (which includes a Solo 401(k)), no Sec.403(a)
// annuity plan, no Sec.403(b) annuity contract -- for substantially the same employees in the
// 3 taxable years before this one. Riley has never run any retirement plan of her own, so that
// lookback is clean. A freelancer coming off her OWN prior Solo 401(k) would fail this test and
// would be limited to the standard \$17,000 figure instead -- the \$18,100 figure is NOT
// automatic just because the employer is small.
const SIMPLE_DEFERRAL_2026 = 18100; // Sec.408(p)(2)(E)(i)(I), per Notice 2025-67 (enhanced small-employer rate)
const matchRate = 0.03; // Sec.408(p)(2)(C)(ii) default 'applicable percentage'
// Sec.408(p)(6)(A)(ii): for a self-employed person, SIMPLE 'compensation' is net earnings from
// self-employment under Sec.1402(a), 'without regard to any contribution under this subsection' --
// i.e. the match base is NOT reduced by the SIMPLE contribution itself (no circularity like some
// other self-employed plans have). That's net SE earnings (92.35% of profit) above, not the
// income-tax 'adjusted earnings base.'
const matchBase = se.netSE;
const ownMatch = matchBase * matchRate;
const ownTotal = SIMPLE_DEFERRAL_2026 + ownMatch;

const SAM_SALARY = 28000;
const samMatchIfDefers3pct = SAM_SALARY * matchRate;
const samNonelective2pct = SAM_SALARY * 0.02;

console.log('--- Option B: SIMPLE IRA ---');
console.log('Riley own elective deferral cap (enhanced small-employer rate):', SIMPLE_DEFERRAL_2026.toFixed(2));
console.log('Riley own 3% match (on net SE earnings, Sec.408(p)(6)(A)(ii)):', ownMatch.toFixed(2));
console.log('Riley own total SIMPLE shelter:', ownTotal.toFixed(2));
console.log('Mandatory cost for Sam -- 3% match design (if Sam defers >=3%):', samMatchIfDefers3pct.toFixed(2));
console.log('Mandatory cost for Sam -- 2% nonelective design (regardless of Sam):', samNonelective2pct.toFixed(2));
console.log();
console.log('Shelter gap (SIMPLE total - Starter total):', (ownTotal - STARTER_2026).toFixed(2));
"

net SE earnings (92.35% of profit): 83115.00
2026 SE tax: 12716.59
deductible half: 6358.30
adjusted earnings base (income-tax AGI concept, NOT the SIMPLE match base): 83641.70

--- Option A: Starter 401(k) ---
Riley own deferral cap: 6000.00
Mandatory employer cost for Sam: 0.00 (statutorily prohibited)

--- Option B: SIMPLE IRA ---
Riley own elective deferral cap (enhanced small-employer rate): 18100.00
Riley own 3% match (on net SE earnings, Sec.408(p)(6)(A)(ii)): 2493.45
Riley own total SIMPLE shelter: 20593.45
Mandatory cost for Sam -- 3% match design (if Sam defers >=3%): 840.00
Mandatory cost for Sam -- 2% nonelective design (regardless of Sam): 560.00

--- Shelter gap (SIMPLE total - Starter total): 14593.45

What that gap is actually worth, after the §199A cap

Riley's graphic-design business isn't a Specified Service Trade or Business, but at $90,000 of net profit she's nowhere near the 2026 single-filer §199A threshold ($201,750, per Rev. Proc. 2025-32) regardless — the SSTB question doesn't bind here either way. What does bind is the separate 20%-of-taxable-income cap under §199A(a)(2), and Treas. Reg. §1.199A-3(b)(1)(vi) treats both plans' deductions as attributable to the trade or business, so both reduce QBI directly:

node -e "
const NET_PROFIT_2026 = 90000, WAGE_BASE_2026 = 184500, STD_DEDUCTION = 16100;
function seTax(p) {
  const netSE = p * 0.9235;
  return netSE > WAGE_BASE_2026 ? (WAGE_BASE_2026*0.124 + netSE*0.029) : netSE*0.153;
}
function bracket2026Single(ti) {
  const b = [[0,12400,0.10,0],[12400,50400,0.12,1240],[50400,105700,0.22,5800],
    [105700,201775,0.24,17966],[201775,256225,0.32,41024],[256225,640600,0.35,58448],
    [640600,Infinity,0.37,192979.25]];
  for (const [lo,hi,rate,base] of b) if (ti > lo && ti <= hi) return {rate, tax: base + (ti-lo)*rate};
}
function scenario(retirementDeduction, label) {
  const half = seTax(NET_PROFIT_2026) / 2;
  const AGI = NET_PROFIT_2026 - half - retirementDeduction;
  const tentativeQBI = AGI * 0.20;
  const taxableBeforeQBI = AGI - STD_DEDUCTION;
  const cap20 = taxableBeforeQBI * 0.20;
  const qbiDeduction = Math.min(tentativeQBI, cap20);
  const taxableFinal = taxableBeforeQBI - qbiDeduction;
  const { rate, tax } = bracket2026Single(taxableFinal);
  console.log('--- ' + label + ' ---');
  console.log('AGI', AGI.toFixed(2), '| tentative QBI', tentativeQBI.toFixed(2), '| 20%-of-TI cap', cap20.toFixed(2));
  console.log('QBI deduction allowed', qbiDeduction.toFixed(2), qbiDeduction === cap20 ? '<- CAP BINDS' : '<- tentative binds');
  console.log('taxable income', taxableFinal.toFixed(2), '| bracket', (rate*100)+'%', '| tax', tax.toFixed(2));
  return { tax, taxableFinal };
}
const starter = scenario(6000, 'Starter 401(k): \$6,000 deduction');
const simple = scenario(20593.45, 'SIMPLE IRA: \$20,593.45 deduction');
console.log();
console.log('Tax saved by using the SIMPLE instead of the Starter 401(k):', (starter.tax - simple.tax).toFixed(2));
"

--- Starter 401(k): $6,000 deduction ---
AGI 77641.70 | tentative QBI 15528.34 | 20%-of-TI cap 12308.34
QBI deduction allowed 12308.34 <- CAP BINDS
taxable income 49233.36 | bracket 12% | tax 5660.00
--- SIMPLE IRA: $20,593.45 deduction ---
AGI 63048.25 | tentative QBI 12609.65 | 20%-of-TI cap 9389.65
QBI deduction allowed 9389.65 <- CAP BINDS
taxable income 37558.60 | bracket 12% | tax 4259.03

Tax saved by using the SIMPLE instead of the Starter 401(k): 1400.97

Four things worth pulling out of that output:

1. The 20%-of-taxable-income cap binds in both scenarios, at a modest $90,000 of profit. This isn't a high-earner problem. In each case the cap (20% of taxable income before QBI) is smaller than the naive 20%-of-QBI figure, so the cap — not SSTB status, not the W-2-wage test — is what actually limits Riley's QBI deduction either way.
2. The $14,593.45 shelter gap doesn't translate dollar-for-dollar into tax savings. Both plans' deductions shrink the same cap they're measured against, so part of each dollar of extra SIMPLE deduction gives back a smaller QBI deduction. The real after-cap value of choosing the SIMPLE over the Starter 401(k) here is $1,400.97, not a naive 12%-of-$14,593.45 estimate (which would suggest roughly $1,751).
3. Both scenarios land in the same 12% bracket, so the comparison doesn't straddle a bracket boundary — the $1,400.97 figure is a clean read, not an artifact of one scenario crossing into a higher rate.
4. None of this is "free" money. The $14,593.45 is still Riley's own cash going into her own retirement account — what she's actually choosing is whether her business can shelter that much of it inside a payroll-deduction plan versus a plain IRA's much lower cap. The real, separate cost of picking the SIMPLE over the Starter 401(k) is the $560 to $840 mandatory contribution for Sam, money that leaves the business regardless of what Riley herself contributes.

If You Have No Employees At All

Strip Sam out of the scenario and the comparison collapses — and not in the Starter 401(k)'s favor. A solo freelancer with no one else on payroll gets a $6,000 deferral cap from a Starter 401(k) for 2026. A plain personal Traditional or Roth IRA, opened with no plan document and no provider relationship at all, carries a $7,500 limit for the same year under §219(b)(1)(A) — $1,500 higher. The two figures started level at $6,000 back in 2024, the Starter 401(k)'s first year, but they run on separate indexing schedules (the Starter 401(k)'s own §402(g)(4)-style formula versus the IRA limit's §219(b)(1) formula) and have already pulled apart; there's no guarantee they'll ever realign. Either way, a Starter 401(k) adds a plan document, a provider relationship, and the mutual-exclusivity trap described above, for a cap that's now smaller than the account you could open this afternoon with no paperwork at all. The plan exists to reach other people's payroll deferrals through auto-enrollment — not to give a one-person business a bigger shelter than an IRA already provides.

Common Mistakes to Avoid

1. Assuming the Starter 401(k)'s cap moves with the regular 401(k) limit. It doesn't — it runs on its own COLA formula starting from a separate $6,000 base, and for 2026 it hasn't increased at all while the regular 401(k) limit rose to $24,500.
2. Assuming a SIMPLE IRA's employer contribution is optional in a lean year. It isn't, short of formally terminating the plan — unlike a SEP-IRA, where skipping a year is simply a discretionary choice.
3. Trying to adopt a Starter 401(k) while already running a SEP, SIMPLE, or Solo 401(k). The "eligible employer" test in §401(k)(16)(E) disqualifies you outright if any of those already exist for the year, and the disqualification runs in both directions.
4. Treating the 2% nonelective and 3% match SIMPLE designs as interchangeable. The 2% nonelective is owed to every eligible employee regardless of their own deferral choice; the 3% match is only owed on top of what an employee actually contributes. Which costs more depends entirely on how many employees actually defer.
5. Opening a Starter 401(k) as a solo freelancer expecting it to beat a personal IRA. For 2026 it doesn't even tie — the IRA limit ($7,500) is actually higher than the Starter 401(k)'s cap ($6,000). The plan's real purpose is reaching other employees through payroll auto-enrollment, not expanding a one-person shelter.
6. Forgetting that both plans' deductions reduce the §199A taxable-income cap. This isn't a high-income-only effect — the worked example above shows it binding at $90,000 of net profit, well below where most freelancers would think to check for it.
7. Assuming the SIMPLE's enhanced $18,100 limit applies just because the employer is small. It doesn't — §408(p)(2)(E)(iv) separately requires that the employer maintained no §401(a)-type plan (including a Solo 401(k)), §403(a) annuity plan, or §403(b) annuity contract for substantially the same employees in the 3 taxable years before adopting the SIMPLE. A freelancer winding down her own prior Solo 401(k) to hire a first employee does not get the enhanced limit on the new SIMPLE — she's stuck with the standard $17,000 figure (or whatever that standard figure is the year she adopts the plan) until 3 full taxable years have passed.

How CentSense Helps

Whichever plan you adopt, the conversation with a SIMPLE IRA provider or a Starter 401(k) TPA starts with an accurate net profit figure — and both plans' contribution math runs directly off that number. CentSense scans every business receipt with AI the moment you take the photo and tracks mileage automatically, so the net-profit figure your retirement-plan decision depends on is accurate and ready well before a provider asks for it. Free tier includes 10 AI receipt scans a month, no credit card required; the Solo plan ($5/month) adds unlimited scans, mileage tracking, and a CPA-ready CSV export.

For the surrounding mechanics this guide assumes, see Solo 401(k) vs. SIMPLE IRA for freelancers hiring their first employee for what happens when an existing Solo 401(k) owner hires someone, SIMPLE IRA vs. SEP IRA for the fully discretionary alternative to a SIMPLE's mandatory contribution, Solo 401(k) contribution limits for the much higher ceiling available to a freelancer with no employees at all, and QBI deduction for freelancers for the taxable-income cap this example ran into twice.

Start free →

Authoritative References

• 26 U.S.C. §401 — Qualified pension, profit-sharing, and stock bonus plans, especially subsection (k)(16), the Starter 401(k) deferral-only arrangement (Cornell LII)
• 26 U.S.C. §408 — Individual retirement accounts, especially subsection (p), SIMPLE retirement accounts (Cornell LII)
• 26 U.S.C. §219 — Retirement savings, especially subsection (g)(5), the cross-referenced "qualified plan" definition that makes the two plan types mutually exclusive (Cornell LII)
• 26 CFR §1.199A-3 — Qualified business income, deductions, and losses, especially subsection (b)(1)(vi) naming self-employed retirement-plan contributions as attributable to a trade or business (Cornell LII)
• IRS Notice 2025-67 — 2026 cost-of-living adjustments for retirement plan dollar limitations, including the Starter 401(k) deferral cap and catch-up amount
• IRS — 401(k) limit increases to $24,500 for 2026, IRA limit increases to $7,500
• IRS Rev. Proc. 2025-32 — 2026 inflation-adjusted standard deduction, tax brackets, and Section 199A threshold amounts

Related reading: Solo 401(k) vs. SIMPLE IRA for freelancers hiring their first employee · SIMPLE IRA vs. SEP IRA · SIMPLE IRA for freelancers · SEP IRA vs. Solo 401(k) · Solo 401(k) contribution limits · QBI deduction for freelancers · 1099 vs. W-2 worker classification

A Starter 401(k) and a SIMPLE IRA are solving different problems — one protects your cash flow, the other maximizes your own shelter — and the statute won't let you run both at once. Start a free CentSense account to keep your net-profit number accurate year-round, so whichever plan you bring to a provider is backed by real numbers, not an April guess. Free tier includes 10 AI scans a month, no credit card required — or upgrade to the Solo plan for $5/month for unlimited scans, mileage tracking, and a CPA-ready CSV export. Start free →

This guide is general education for U.S. self-employed freelancers filing in 2026. It is not personalized tax or retirement-planning advice. Which plan is actually available to you, how matching contributions are computed for a self-employed owner, and how either deduction interacts with your full tax return are fact-specific determinations. Consult a CPA, EA, or retirement-plan provider before adopting either plan.