The price of AI consulting is hard to pin down because most firms don't publish one. The guides that rank for it mostly give a range so wide it can't be budgeted, from a few thousand dollars to half a million. This guide uses only prices we could see in October 2026: rates in a US government database, and fixed fees that firms put on their own pages. They show a market that's cheaper and more predictable at the small end than the guides suggest.
How much does AI consulting cost?
As of October 2026, firms that publish prices charge $4,000 to $12,500 for a fixed-fee AI readiness assessment and $1,450 to $15,000 for a strategy workshop, AI work billed by the hour has a median ceiling of $187.64 an hour on US federal contracts, and AI development projects reviewed on Clutch most often cost $10,000 to $49,999.
This guide is for the people who sign off the spend at a company of 20 to 1,000 people: COOs, finance leads and IT leads who've been sent a proposal, or who want a figure before they ask for one. It covers what each common engagement costs, what you get for the money, why quotes for the same job differ, and how to budget.
Two things to know about the numbers. Every price here comes from a page we read in October 2026, and the Sources section lists each one. Most firms price by scope, so a published price is a guide to the market, not a quote for your project. And Deployment Labs sells consulting too. We don't list our own prices in this guide, and we've left them out of every range.
| Type of work | Published price | Typical length |
|---|---|---|
| Hourly or day-rate advice | Median ceiling $187.64 an hour | As needed |
| AI readiness assessment | $4,000 to $12,500, fixed | 1 to 3 weeks |
| Strategy workshop | $1,450 to $15,000, fixed | Half a day to 2 days |
| Implementation project | Most often $10,000 to $49,999 | Months |
| Staff training | $199 a user a year (self-paced) | Ongoing |
| Advisory retainer | $3,000 to $5,000 a month (one firm) | Monthly |
What do AI consultants charge per hour?
On US federal contracts, the median ceiling rate for AI and machine learning work is $187.64 an hour, and half of all rates fall between $148.83 and $243.74.
That figure comes from CALC+, the US General Services Administration's database of hourly rates awarded on its Multiple Award Schedule. We pulled every AI and machine learning labour category in it on 25 October 2026: 251 rates from 81 firms, for roles from junior engineer to subject matter expert.
Read these rates as ceilings. GSA's own developers explain that "contract-awarded prices are never actually what anyone pays for a labor service." They are "essentially the highest price ("ceiling") that could be paid to a vendor", and "in practice there are always discounts" negotiated for each order.
Three patterns in the data are useful when you read a proposal:
- Seniority moves the price most. Titles marked junior have a median of $117.33 an hour, senior $217.98 and principal $284.26, though those are small groups of 8, 15 and 5 rates. Only 20 of the 251 rates are $300 an hour or more, and the highest is $485.47.
- Advice costs a little more than the average. Roles titled consultant, strategist or subject matter expert have a median of $202.15 an hour across 34 rates, against $187.64 for all 251.
- Smaller firms charge less. Rates from small businesses have a median of $178.00 an hour, against $220.54 from larger firms.
Commercial rates spread wider. Clutch, a review site, says most AI development companies it lists charge under $50 an hour. Those are mostly software development companies rather than advisers, and the figure covers the whole list. At the other end, one US boutique publishes a standard rate of $175 to $225 an hour "depending on scope and compliance requirements", which sits close to the federal figures.
What does an AI readiness assessment cost?
Published fixed fees for an AI readiness assessment run from $4,000 to $12,500 for one to three weeks of work, as of October 2026.
An assessment is a common first purchase, and the kind of AI consulting we found with a public price most often. It looks at the AI your company already uses, your data, your rules and your people, and ends with a list of what to do first.
| Firm's offer | Published price | Length | What you get |
|---|---|---|---|
| Avantiico, AI Readiness Assessment | $4,000 fixed | About 1 to 2 weeks | Scorecard on five dimensions, gap analysis, recommendations |
| AIgist24, AI assessment | $4,999 | 2 weeks | Scorecard, report, 3 to 5 use cases, roadmap |
| Aries Consulting, AI readiness audit | $7,500 fixed | 2 to 3 weeks | Scorecard, stack audit, 90-day roadmap |
| Veracity AI, governance readiness assessment | $8,000 to $12,000 fixed | 2 to 3 weeks | Governance review |
| WorkWise Solutions, AI Readiness Sprint | $12,500 fixed | Not stated | Assessment, credited against later work |
Two details in these offers are worth asking any firm about. WorkWise says its fee "credits in full toward any engagement of $25,000+ booked within 90 days", and one boutique says its discovery work, from $3,500, applies toward a later build. Aries says its audit is "decoupled from any follow-on build". Both approaches are reasonable. What matters is knowing which one you're buying, because a credited assessment is partly a sales step.
Much of the work in these offers is the same groundwork: interviews and surveys to find out which AI tools staff already use. You can do a first pass yourself in minutes with our free Shadow AI Audit, or our AI Readiness Assessment, and arrive with the list instead of paying for it.
What does an AI strategy workshop cost?
A facilitated AI strategy workshop is published at $1,450 to $15,000, depending mostly on length and on who it's for.
- An independent consultant in Florida publishes a half-day workshop for up to 12 people at $1,450 remote or $1,950 on-site, per organisation.
- Kriv AI lists an executive strategy workshop on AWS Marketplace, aimed at regulated companies with $500 million to $5 billion in revenue: $5,000 for a half day, $10,000 for a full day and $15,000 for a two-day offsite, with travel billed separately.
The difference is what surrounds the room. The $1,950 workshop is four hours with your team. Kriv AI's includes interviews beforehand and, within 10 business days afterwards, a strategy deck, an executive memo and a reference architecture. Price the follow-up, not the hours.
What does AI implementation cost?
AI development projects reviewed by verified clients on Clutch most often cost $10,000 to $49,999, with an average of $120,594.55 and a typical timeline of 10 months.
Implementation is where the range opens up, because "implementation" covers everything from setting up one workflow in software you already pay for to building a custom system. The average is pulled up by a few large projects. The most common bracket is the better guide for a company of a few hundred people.
Three things decide where your project lands:
- How much is new. Switching on and configuring an AI feature inside Microsoft 365, Google Workspace or your CRM costs a fraction of building something that connects several systems.
- How many workflows it touches. One workflow, proved and then widened, costs less and fails cheaper than a platform for the whole company.
- What you pay to run it. The consulting fee is a one-off. The model, licence and usage fees are not, and they belong in the same budget.
What does AI training cost per employee?
Self-paced AI courses can cost about $199 a user a year, while live workshops are priced per team, from about $1,450 for a half day.
One provider publishes its AI course library at "$199/year per user, volume discounts at 30+ seats". For live training, the half-day workshops above are the published benchmark. Many companies use both: a short live session on their own work, then a course for reference.
Free options are good enough for the basics. Our Everyone's first hour with AI course is free for all staff, and if you want training built around your own tools and rules, see our employee AI training.
Do AI consultants charge a retainer?
Some do, for ongoing advice after the first project, and published retainers start at a few thousand dollars a month.
Veracity AI publishes an advisory retainer of $3,000 to $5,000 a month for ongoing governance review. Other firms sell a block of hours each month at their standard rate, so a retainer of 20 hours at $175 an hour would be $3,500 a month.
A retainer suits you once something is running and needs an owner's second opinion. It suits you less before you've decided what to build, when it can become paid open-ended exploring. Ask for the hours or days included, what happens to unused time, and how much notice ends it.
Why do quotes for the same AI project differ so much?
Because they're rarely quotes for the same project. The scope, the seniority of the people, the systems involved and the compliance work all change the price, and the published offers above show how much the deliverables vary under the same name. Two proposals can share a title and differ several times over because one includes the integration work and one leaves it to your IT team.
The fix is to decide what you're buying before you ask. In our framework, every piece of work that costs money gets a one-page case first: what it frees at the bottleneck, what it costs in full (fees, licences and your own people's time), who owns it, and the result that would make you stop. Our free lesson on the one-page case shows how to write it, and the lesson on spending limits covers capping what runs after.
With that page in hand, quotes become comparable. Send the same page to two or three firms, ask each to price exactly that, and the differences that remain are about quality and approach, not scope.
How to budget for AI consulting
01 · List what's already in use
Find the AI tools and features your company already pays for and uses, so no one bills you for discovering them. This gives you the starting point for every quote.
02 · Name the one problem worth paying to solve
Pick the bottleneck where work waits longest, and put a value on clearing it. Our lesson on naming the bottleneck shows how. That value is the ceiling on what the work is worth.
03 · Write the one-page case
State the result, the owner, the budget, the running costs and the stop condition. This is the scope you'll ask firms to price.
04 · Ask for fixed fees against the same page
Ask two or three firms to price the page, with deliverables and dates. Ask whether an assessment fee is credited against later work, what's excluded, and what the tools will cost to run after the work ends.
05 · Buy the smallest proof first
Start with an assessment or a single workflow that proves itself in weeks, then widen. A small proof that fails costs a few thousand dollars. A large build that fails costs the year.
Deploying AI is mostly a question of picking the right work and owning it once it runs. If you'd like help writing the case and running the first proof, that's what our AI implementation team does, starting from a consultation.
What's still open
- Prices are moving. Firms reprice AI work often, and several of the pages here were updated within the last few months. Treat every figure as "as of October 2026".
- Only published prices are here. Firms that only quote case by case aren't in these ranges, so the top of the market is probably missing.
- Federal rates are ceilings. CALC+ shows the most a firm may charge on a schedule contract, not what was paid on a given order, and the rates change as contracts are renewed.
We'll update this article when published rates move materially or a larger set of prices becomes public. The date at the top of the page shows when we last checked.
Sources
- CALC+ ceiling rates API, US General Services Administration, October 2026: the 251 AI and machine learning labour rates from 81 firms on the Multiple Award Schedule, downloaded on 25 October 2026. The median, quartiles and splits are our own calculation.
- CALC README, GSA 18F, GitHub: what the awarded rates mean, and that orders are discounted below them.
- AI Pricing Guide, Clutch, September 2026: hourly rates of listed AI development companies, and project costs and timelines from verified client reviews.
- AI Readiness Assessment Services, Avantiico, October 2026: a published fixed-fee assessment.
- AI Assessment Service, AIgist24, October 2026: a published assessment price and deliverables.
- AI Readiness Audit Cost, Aries Consulting Group, April 2026: a published fixed-fee audit and its deliverables.
- What an AI governance assessment costs, Veracity AI, June 2026: a published assessment fee and advisory retainer.
- What does AI readiness cost?, WorkWise Solutions, October 2026: a published sprint fee and its credit terms.
- AI Consulting Rates and Pricing in 2026, Layer3 Labs, September 2026: a boutique's published hourly rate, discovery price and hour-bank retainers.
- Executive AI Strategy Workshop for Regulated Mid-Market Enterprises, Kriv AI on AWS Marketplace, October 2026: published workshop prices and what follows them.
- AI workshop booking page, Cal.com, October 2026: an independent consultant's published workshop prices.
- AI Training for Employees, Iternal, July 2026: a published per-user price for a self-paced AI course library.