Free tool

What does a developer really cost you?

Salary is only the start. This calculator adds on-costs, recruitment, months spent hiring and ramping up, attrition, management time and AI tooling, then compares five ways to get the same engineering output, including a senior team that works AI-first.

Calculator

Price your engineering capacity

Choose a starting point, then replace salaries and rates with figures from your own market.

01Where would the engineers be based?

Changing location resets salary, on-costs and rates to typical figures for that market.

02Seniority
03How much capacity, for how long?
4
04In-house employment cost
$USD / year
17%
20%
05Hiring and retention

Time to hire applies to in-house roles. Ramp-up and attrition apply to every model, scaled by how each one works.

2.5 months
3.0 months
15%
06Overheads and tools
12%
80%
$USD / engineer / yr
$USD / engineer / mo
07External rates

Seeded from local salaries. Overwrite them with real quotes; all figures are USD per hour.

$USD / hour
$USD / hour
$USD / hour
$USD / hour
08AI productivity assumption

The uplift a senior engineer gets from working AI-first. Each model realises a different share of it, and juniors benefit less reliably.

25%
True cost of capacity
Best value for this need
Close call
Freelance and AI-native are within 10% at about $17,000 per productive engineer-month, so weigh speed, control and continuity.
In-house / year
$380,000
Best value / year
$470,000
Fastest start
AI-native, 0.8 mo
AI-native headcount
2.5 vs 4
Cost per productive engineer-month
  • In-house hire$19,000
  • Freelancers$17,000
  • Development agency$19,000
  • Staff augmentation$18,000
  • AI-native senior team$18,000
Cost per year
  • In-house hire$380,000
  • Freelancers$470,000
  • Development agency$540,000
  • Staff augmentation$470,000
  • AI-native senior team$510,000
Time to first output
  • In-house hire3.3 months
  • Freelancers1.2 months
  • Development agency1.4 months
  • Staff augmentation1.6 months
  • AI-native senior team0.8 months
Recommendations
  • Freelancers look cheapest, but coordinating four or more independents takes real management time and consistent standards. An agency or augmentation partner adds structure.
  • The lowest cost per productive engineer-month here comes from the freelance option, at $17,000 versus $19,000 for in-house.
  • In-house hire has the lowest bill ($380,000 a year) but loses more of the period to hiring, ramp-up or turnover, so each productive month costs more.
AI impact
  • An AI-native senior team matches this output with about 2.5 people instead of 4, at $18,000 per productive engineer-month, assuming a 25% uplift for engineers who work AI-first.
  • Equipping in-house hires to work AI-first (tools, training and review conventions) would bring their cost per productive month to about $17,000 under the same assumption.

A productive engineer-month is one month of focused work from a senior engineer without AI tooling. External models bill 140 hours a month. Indicative 2026 assumptions, not salary survey data.

Indicative only. We will send the inputs above with your message so a senior engineer can sanity-check them.

The cheapest hourly rate is rarely the cheapest engineering output.

A salary looks cheaper than an agency rate until you count the three months of searching, the months before a new hire is fully productive, the recruiter invoice, the laptop, the licences and the share of a manager's week. A low freelance rate looks cheaper still until two of your four freelancers move on mid-project.

That is why this calculator measures cost per productive engineer-month: one month of focused senior-level work. It rewards models that start quickly, keep people and need little management, and it penalises the ones that look efficient on an invoice but lose weeks to friction.

Hidden costs

What the true cost of hiring a software developer includes

Every card below is an input in the calculator. Most hiring budgets capture the first one and guess the rest.

Benefits and on-costs

Employer pension or retirement contributions, insurance, payroll taxes and leave typically add a fifth to a third on top of base salary, depending on the country.

Recruitment and time to hire

Agency fees are commonly 15 to 25% of first-year salary, and every month the seat stays empty is a month of roadmap that does not ship.

Ramp-up

New engineers need time to learn the codebase, the domain and the team. Expect reduced output for the first few months, even from strong hires.

Attrition

Each departure restarts the cycle: another search, another fee, another ramp-up, and knowledge that leaves with the person.

Management and utilisation

Someone plans, reviews and unblocks the work. Meanwhile meetings and interruptions mean no engineer spends every paid hour building.

AI tooling and ways of working

AI coding tool seats are a small line item. The larger investment is training engineers to work AI-first and setting review standards for generated code.

Rules of thumb

Typical ranges to sanity-check your inputs

20 to 35%
On-costs over base salary

Varies widely by country and benefits policy

15 to 25%
Recruitment agency fee

Of first-year base salary, for permanent roles

2 to 4 months
Time to hire

From opening a role to a signed offer and start date

3 to 6 months
To full productivity

Longer for complex domains or legacy systems

Indicative ranges commonly seen in the market, not survey results. Use your own figures where you have them.

Hiring in the AI era

AI-native engineers change the headcount question.

Engineers who work AI-first run coding agents on routine code, tests and documentation, then spend their own time on design, review and the hard problems. A small senior team working this way can often match the output of a larger traditional team, which is why the calculator sizes the AI-native option to your required output rather than your headcount.

The uplift is an assumption you control, and it deserves scepticism. It varies with the codebase, the type of work and, above all, the engineer. Juniors benefit less reliably because they are less able to spot where generated code is subtly wrong. The model reflects that by giving seniority a larger share of the AI benefit.

If you hire in-house, test for this directly. Give candidates a realistic task with AI tools available and watch how they break the problem down, what they ask the model for, and how they verify the result. Budget for training and review conventions, not just tool seats, or the uplift will stay theoretical.

FAQ

Questions we often hear

How much does it cost to hire a software developer?

The fully loaded cost is usually well above base salary. Once you add on-costs, recruitment fees, equipment, tools and management time, first-year cost is often 1.4 to 1.8 times salary, and more if the role takes months to fill or the hire leaves early. The calculator shows the figure for your market and assumptions.

Is outsourcing cheaper than hiring in-house?

For short or uncertain needs, outsourcing is often cheaper because you avoid recruitment fees, empty seats and ramp-up. For a stable team over two years or more, in-house hiring usually earns back those upfront costs. Many companies combine the two, using an external team while they hire.

What is the difference between staff augmentation and an agency?

With staff augmentation, a provider supplies engineers who join your team and work under your management and processes. An agency delivers an outcome with its own project management, QA and team structure. Augmentation is usually cheaper per hour but needs more of your management time.

Do AI coding tools mean I need fewer developers?

Often you need fewer people for the same routine output, especially with senior engineers who work AI-first. You do not need fewer senior decisions: architecture, security, product judgement and code review still take experienced people. Treat any productivity figure as an assumption to test, not a promise.

How long does it take a new developer to become productive?

Most new engineers make useful contributions within their first few weeks but take three to six months to reach full productivity. Clear documentation, a working local setup, a patient mentor and AI tools that help explain unfamiliar code all shorten that period.

Build the right team

Deciding whether to hire or partner?

Send us your results and what you need to build. We will tell you honestly whether to hire, which roles to open first, or whether an AI-native team gets you there faster, and we can help screen and interview candidates if you hire.

Working with companies globally · Response within 24 hours