Tier 1: AI assistants (from free to about $20 a month per person)
General assistants handle writing, summarizing, brainstorming and analysis. Each of the big three has a free tier and a paid individual plan. Prices checked October 2026 on each vendor's own pages:
| Tool | Paid individual plan | Source |
|---|---|---|
| ChatGPT Plus | $20/month, billed monthly | OpenAI Help Center |
| Claude Pro | $20/month billed monthly, or $17/month billed annually | Anthropic pricing |
| Google AI Pro (Gemini) | $19.99/month | Gemini subscriptions |
All three also sell business or team plans priced per user, with admin controls and different data terms. Check the vendor pages for current business pricing, because prices and plans change. For which one to choose, see ChatGPT vs. Claude vs. Gemini for business.
Also check software you already pay for. Accounting, field-service, CRM and email tools increasingly include AI features in existing plans or as paid add-ons, and that may be your cheapest option.
Tier 2: done-for-you AI tools (monthly fees that vary with usage)
These are products built for one job: AI receptionists, missed-call text-back, review management, scheduling, chat widgets, bookkeeping helpers. Prices vary too much between vendors, and change too often, for a fair range here. Instead, compare quotes on these cost drivers:
- Usage: per minute, per call, per message, per contact or per seat. Price it against your busiest month, not your average.
- Integrations: connecting to your calendar, CRM or field software often sits on a higher plan.
- Setup and onboarding fees, and whether you or the vendor does the configuring.
- Contract length and minimums, and what leaving costs.
For phones specifically, our AI receptionist guide shows how to work out whether a tool pays for itself.
Tier 3: custom builds (priced per project)
A custom build means someone connects your systems and builds a workflow for your business: for example, quote requests that flow from your website into your CRM with an AI-drafted reply, or a tool that answers staff questions from your own documents. Quotes vary a lot. These are the main drivers:
- Number of systems connected and whether they have usable integrations.
- State of your data. Messy spreadsheets and duplicate customer records add cleanup time.
- Number of workflows and how many exceptions each one has.
- Who maintains it. AI models, APIs and your own processes change, so ask what ongoing support costs.
- Usage costs for the AI models underneath, which are usually billed separately by usage.
Get the scope in writing, ask what happens when an underlying tool changes, and make sure you own your accounts and data.
The hidden costs
- Your time. Setup, writing prompts and templates, and checking output in the first weeks. This is usually the biggest cost.
- Training your team to use the tool the same way. See how to train employees to use AI.
- Seat creep. Per-person pricing grows quietly as you add staff.
- Overlapping subscriptions. Three tools doing the same job is common.
- Mistakes. An unchecked AI message to a customer has a cost too, which is why every guide here keeps a person in the loop.
How to budget without overspending
- Start with Tier 1 and the AI features in software you already pay for.
- Pick your first task with our scoring sheet and run a two-week test.
- Move to a Tier 2 tool only when a test shows the task is worth paying for.
- Consider Tier 3 only once the process is stable and you know exactly what you want built.
My AI Plan costs $497, one fixed price. It's a planning step that comes before the bigger spending decisions. From a 15 to 30 minute interview, you get a written plan that names specific tools for your business, each with a monthly cost and setup time, plus a 30-minute expert review call. So you know what to buy, and what not to, before you commit to a tool or a build.