How to price AI automation services: setup fee, retainer and margin
Most AI automation pricing advice is either a single magic number or a range so wide it says nothing. The useful answer has three parts: what the market has been observed to charge, what the work really costs you, and what the result is worth to the client. This guide covers all three, for an agency with a roster and for a founder quoting a first client.
Direct answer
How should I price AI automation services?
Charge a one-time setup fee for discovery and the build, plus a monthly retainer for running, watching and improving the automation. Set the floor from your cost: price = cost divided by (1 minus your target margin), so $600 of build time at a 70% margin is a $2,000 setup fee. Count every monthly cost in the retainer: your share of the platform plan, the client's hosting and model usage, and your own hours. Then check the price against the value to the client and against observed market ranges, which for small businesses cluster around low thousands for setup and hundreds to about $1,500 a month, as of September 2026.
Observed AI automation pricing, as published
Ranges as each page states them, in US dollars, checked on 30 September 2026. Most are guides written by agencies or vendors without a stated method; treat them as observed quotes, not earnings.
| Source | Client or scope | Setup or project | Monthly | Basis |
|---|---|---|---|---|
| Taskip | Small business | $1,500 to $5,000 | $500 to $1,500 | Compiled from other guides |
| Taskip | Mid-market | $5,000 to $20,000 | $1,500 to $4,000 | Compiled from other guides |
| Evolv AI Agents | Lead follow-up system | $2,000 to $8,000 | $400 to $1,500 | Public rate cards, no sample size |
| Layer3 Labs | US agency engagements | $5,000 to $75,000 | $3,000 to $20,000 | Stated typical ranges |
| Parix.ai | Simple builds | $1,500 to $4,000 | 5 to 15% of setup per year | Its own projects |
| Goodspeed | n8n workflows | $5,000 for up to 4 workflows | From $10,000 | Its own published prices |
Start with the structure: a setup fee plus a monthly retainer
The setup fee pays for the work before anything runs: discovery, access and credentials, building and tuning the automation, and handing it over. The retainer pays for everything after: the running cost, watching it, fixing what breaks when a client's tools change, and small improvements. Keeping them separate protects you twice. A client who leaves in month two has still paid for the build, and a retainer that stands on its own makes the monthly value visible at renewal. Package the retainer as a named service with a clear scope, such as lead follow-up for one inbox, rather than hours, so the price reads as an outcome.
What AI automation agencies charge: observed ranges
Few agencies publish their prices, so most figures online come from pricing guides written by agencies and software vendors, usually without a stated method. Read them as the range of what is quoted, not as a promise of what you will earn. Every figure below was checked on 30 September 2026 on the page cited. They disagree by an order of magnitude, mostly because they describe different clients: a single workflow for a local business is priced in the hundreds a month, a managed program for a mid-sized company in the thousands.
- Taskip (updated 18 August 2026): small business setup $1,500 to $5,000 and retainers $500 to $1,500 a month; mid-market setup $5,000 to $20,000 and $1,500 to $4,000 a month. Compiled from other 2026 pricing guides.
- Evolv AI Agents (2026): setup $1,500 to $15,000 and monthly management $300 to $5,000 overall; lead follow-up systems $2,000 to $8,000 setup and $400 to $1,500 a month. Compiled from public rate cards and industry surveys, with no sample size.
- Layer3 Labs (2 July 2026): US agencies at $5,000 to $75,000 per project and $3,000 to $20,000 a month on retainer, and $5,000 to $15,000 for a small first project. It notes that almost no agency publishes these numbers.
- Parix.ai (5 June 2026), from its own projects: simple builds at $1,500 to $4,000, with running costs of 5 to 15% of the setup price per year.
- Goodspeed, an n8n partner, publishes its own prices: $5,000 fixed for up to four production workflows, and an ongoing automation team from $10,000 a month.
- Clutch (September 2026), from verified client reviews of AI development companies: most listed firms charge under $50 an hour, and the most common project size is $10,000 to $49,999.
Sources: Taskip, AI automation agency pricing, checked 30 September 2026 · Evolv AI Agents, AI automation agency cost, checked 30 September 2026 · Layer3 Labs, AI automation agency cost, checked 30 September 2026 · Parix.ai, how much AI automation costs, checked 30 September 2026 · Goodspeed, n8n agency pricing, checked 30 September 2026 · Clutch, AI development pricing, checked 30 September 2026
Know your cost first: platform, model usage and your time
Your cost has four lines, and the one people forget is their own time. First, the platform plan, shared across clients: on Qoren, Ultimate costs $169 a month for up to 10 environments, so a full roster puts $16.90 of plan fee on each client. Second, the client's hosting, billed by the hour for their environment. Third, usage: model calls, web searches and a mailbox if the agent has one. Model usage is the line that moves most from month to month. Fourth, your hours: the build, then a monthly check-in, tuning and the occasional fix. You do not have to guess the platform lines after the first month, because the client's Cost tab shows what their environments and agents used, split into hosting, models, web search and mailboxes.
Cost-plus sets the floor, value sets the ceiling
Cost-plus pricing gives you the lowest price you should accept: price = cost divided by (1 minus your target margin). A worked example, with assumptions you should replace with your own: 8 hours of build at $75 an hour is $600, which at a 70% margin is a $2,000 setup fee. Each month, the client carries $16.90 of plan fee, uses $60 of hosting and model usage on its Cost tab, and takes 1.5 hours of your time ($112.50). That is $189.40 a month, so the retainer at a 70% margin is $640 once you round up to the next $10, which leaves the margin at 70%. Value pricing asks the other question: what is the result worth? If the agent answers every lead within minutes and one extra booked job a month is worth more than the retainer to the client, price toward the value and keep the cost-plus figure as your floor.
- A 70% margin is not a 70% markup: the price is 3.33 times the cost, a 233% markup. Quote margins, and check which one a tool shows you.
- Your first client carries your whole plan. On Starter ($39 a month), the same example costs $211.50 and prices at $710; the share falls as you add clients.
- Proposals passes the plan fee through at cost, with no margin on it, so its suggested price sits slightly below a hand calculation that marks everything up. Pick one approach and use it for every client.
Price the pilot, do not give it away
A pilot lowers the client's risk, not your price. Offer one agent for one job, for two to four weeks, at a fixed fee, with the success measure written down before it starts, such as the share of leads answered within ten minutes. Credit the pilot fee against the setup fee if the client continues, so saying yes feels like progress rather than a second purchase. Run the pilot in the client's own environment from day one: if they sign, nothing moves, and the pilot's real usage tells you what the retainer has to cover. A free pilot teaches the client that the work is worth nothing and gives you no cost data you can trust.
Quote it, check it against real cost, and raise it
Build the quote in Proposals, inside the console. Add the agents the client needs, and it prices hosting and usage from Qoren's live rates; you set the target margin and setup fees, and it shows your cost and profit on every line while the client document shows only prices. When you cannot predict model usage, enter a low and a high estimate and the proposal quotes a range. After the first full month, open the client's Cost tab and compare it with the proposal: real usage comes in above or below the estimate, and that gap is your first pricing signal. Raise prices at renewal, when you add a workflow, or when a client's usage has moved for good. Give notice, tie the change to results the client has seen, and reprice new clients first.
Related guides
For agencies and resellers
You sell the agents. We do the homework.
A platform for agencies to build, deploy and resell managed AI agents to SMB clients: an isolated environment per client, one console, your own invoice.
Read guidePlaybook
How to start an AI automation agency, and sell what SMBs actually buy
One guide to starting an AI automation agency: what small businesses actually buy, how to deliver it with managed agents, and real retainer pricing.
Read guideAgency operations
How to run AI agents for multiple clients without mixing them up
How to run AI agents for many clients: an isolated environment and credentials per client, grouping by client, one account budget and cost per client.
Read guideTools, templates and next steps
Pricing calculator
Turn your costs and target margin into a setup fee, retainer, and MRR.
OpenProposal generator
Pick automations, set pricing, and copy a client-ready proposal.
OpenWrite a client proposal for AI agents
Quote a client for AI agents run on Qoren with your platform cost and margin in view, then print a proposal under your own brand or save it as a PDF.
OpenSee what each client costs
See what each agency client cost you on Qoren over a period, split into hosting, models, web search and mailboxes, so you can bill your clients with confidence.
OpenAI automation retainer
An AI automation retainer is a recurring monthly fee a client pays an agency to keep its AI agents running, maintained, and improved after the initial build.
OpenQoren pricing
Every plan, with what each one includes.
OpenFrequently asked questions
How much should I charge for AI automation services?
Start from your cost, not from a market average: price = cost divided by (1 minus margin). In this guide's example that is a $2,000 setup fee and a $640 monthly retainer at a 70% margin. Published guides checked on 30 September 2026 put small business work at roughly $1,500 to $5,000 for setup and $500 to $1,500 a month, with mid-market work well above that.
Should I charge a setup fee or only a monthly fee?
Charge both. The setup fee pays for discovery and the build, which you do whether or not the client stays. The retainer pays for running and improving the automation. Folding setup into the monthly price only works with a minimum term long enough to recover the build.
Is cost-plus or value pricing better for AI automation?
Use both. Cost-plus gives you the floor, the lowest price that keeps your target margin. Value pricing gives you the ceiling, what the result is worth to the client. Quote between them, closer to value when the outcome is measurable.
How do I price model usage that changes every month?
Estimate it as a range and price the retainer on the high end, or quote it as a range. Proposals can price each agent's model spend at a low and a high estimate and show the client a range. After the first month, the client's Cost tab shows what the usage really was.
How do I raise prices for existing clients?
At renewal, when you add scope, or when their usage has changed for good. Give notice, show what the automation delivered, and apply the new price to new clients first so you know the market accepts it.
Run every client's agents from one console.
Give each client an isolated environment with its own credentials, deliver the work under your name or done-for-you, and keep the whole account inside one budget.