What AI consulting actually costs
Real bands for assessments, builds, retainers and training, what moves each number, and why almost nobody in this market will tell you before a sales call.
An AI opportunity assessment runs roughly $15,000 to $25,000 as a fixed fee. A build starts around $40,000 and most land between $40,000 and $120,000. A fractional AI advisory retainer sits between $4,000 and $8,000 a month and a fractional CTO between $6,000 and $15,000. One day of training is around $4,000. Those are our published bands. The wider market mostly refuses to publish anything, so a buyer comparing three firms is usually comparing three sales calls rather than three prices.
When we read seven AI consultancies' own websites in September 2026, six of them published no price at all. Not a range, not a starting figure, nothing. The seventh explained what moved its pricing without naming a number.
That is the market a buyer walks into. You cannot shortlist on budget, because budget is the one thing nobody will tell you until you have spent half an hour on a call.
So here are the bands, with what moves each one.
The assessment
$15,000 to $25,000, fixed fee.
Two to three weeks. Somebody sits with the people doing the work, finds where the hours actually go, and comes back with the short list worth building, the longer list that is not, and a baseline figure for each survivor.
What moves it is how many teams have to be interviewed and how quickly access to systems arrives. A single department in one company sits at the bottom. Four departments across two entities sits at the top.
The thing that should not move it is the size of the build that follows, and if a firm's assessment fee scales with the project it recommends, that is worth noticing.
The build
From $40,000. Most land between $40,000 and $120,000.
What moves it is not the model. It is how many systems the thing has to touch, whether those systems have usable interfaces, and how much of the work is making other people's software cooperate.
A single automation into a stack with clean interfaces lands near the bottom. A feature inside a product, with evaluation, monitoring and a rollback path, lands in the middle. Anything touching a system nobody has modified in six years lands at the top, and the reason is archaeology rather than engineering.
Ask any firm quoting a build what proportion of the estimate is integration. If they cannot answer, the estimate is a guess.
The retainer
$4,000 to $8,000 a month for advisory. $6,000 to $15,000 for a fractional CTO.
The difference is accountability. Advisory means somebody senior in the room for the decisions. Fractional CTO means somebody owns the outcome, runs the engineers and answers for what ships.
Both of ours run on one month's notice, which is unusual enough to be worth asking about elsewhere. A twelve-month minimum on a retainer is a firm protecting its revenue rather than your outcome.
The training
$4,000 for a half day. $30,000 for a four-week programme.
What moves it is whether the day is taught against your team's own work or against a curriculum. A curriculum is cheaper to deliver and is the thing most platforms sell. Working on the team's real backlog costs more and is the version that changes anything.
The small one
$2,500 fixed, for one workflow in seven days.
This exists because the assessment is the right first purchase for a company with a real budget and the wrong one for a company that wants to find out whether any of this works. Seven days, one job, a working system and the before and after figures.
It is worth doing when a single job takes a team ten hours a week or more between them. Below that the arithmetic does not work, and we will say so rather than sell it.
What makes a quote hard to compare
Three things, and they are the reason two quotes for the same project can differ threefold.
What counts as done. One firm's estimate ends at a working demo. Another's ends with the thing in production, monitored, documented, and handed to your team. These are not the same project and they should not be the same price.
Who does the work. A quote priced on a senior name and delivered by whoever is free is the oldest problem in consulting. Ask who specifically, and ask how often you will see them.
What happens when it is wrong. Most estimates assume the first approach works. The honest ones carry a line for the version where it does not.
One limit worth stating
Our own bands were derived rather than confirmed. They were set in August 2026 from market research and from our build-cost tool's own model, and they have not yet been corrected against a full year of invoices. The ones most likely to be wrong are builds, which hang off an assumed blended weekly rate, and training, where the market range puts us mid-market and that may be below where we should sit. They are published anyway, because a derived number you can argue with beats a number withheld until a sales call.
Before you ask anyone for a quote
A brief that is specific enough to compare answers is worth more than another conversation. How to brief an AI project so the quotes are comparable sets out the five things to write down first.
If you would rather see the shape of the market than the arithmetic, seven firms are compared here, including what each one does and does not publish. Our own figures are on the pricing page, and the call is free.
If you have a budget, a deadline, and no clear answer on which AI project deserves either, that is the conversation to have.
Book the callWritten by
Radwan Altaf
Radwan runs AISynq. Before that he delivered software inside enterprise programmes at DHL, AT&T, DirecTV and Accenture, which is where the habit of measuring a result against its baseline came from. More about the firm.
Read next
- The AI project brief template: fill this in before you ask for a quoteA one-page template with the nine fields that make three quotes comparable, what to write in each, and the two answers that are allowed to be "we do not know".
- The AI vendor scorecard: compare three firms on one pageA scoring template for comparing AI consultancies on the twelve things that predict whether an engagement works, with what each score means.
- How to brief an AI project so the quotes are comparableFive things to write down before you contact anyone, and why a vague brief produces three quotes that differ threefold for reasons nobody can explain.
Work with us
Have a budget, a deadline, and no clear answer on which AI project deserves either?
Book the callGet the next one
New writing in Buying AI help as it goes up, roughly twice a month. One article per email and nothing else in it.