How to price and rebill GoHighLevel AI usage to clients

By / Published Nov 12, 2026 / Updated Nov 12, 2026

Heads up: the trial and plan buttons on this page use my affiliate links. If you sign up through one I get a commission. It doesn't cost you more.

To rebill GoHighLevel AI to a client, turn on rebilling for their sub-account (Agency View > Sub-Accounts > Manage Client > Rebilling) and pick a pricing model for Conversation AI and for Voice AI separately: a markup on HighLevel's token cost, or a fixed rate per message or per minute that you set. HighLevel always charges your agency wallet the real cost first, rebilling AI usage needs the $497 Founder plan, and nothing stops a usage spike until you set a limit to Block AI (as of Nov 12, 2026).

This page is the margin side only: which model to pick, where each one can lose you money, how to cap spikes and agent loops, and how to show a client what they used. Every per-minute and per-token rate is on Usage fees, and what each AI Employee plan includes is on GoHighLevel AI tools.

Start your 30-day free trial →

30 days free instead of HighLevel's usual 14, plus a 1:1 setup call with my team, weekly live calls and my agency resources from day one. See everything you get →

Short version

  • 01Four ways to charge: fold AI into your flat monthly fee, pass it through at cost, rebill with a markup, or rebill at a fixed rate per message or minute. Or resell an AI Employee plan at a price you set.
  • 02Conversation AI and Voice AI have been separate billing products since Jul 30, 2026. Set each one, because changing one doesn't touch the other.
  • 03Fixed rate is the easiest to sell and the only one where you can lose money per unit. HighLevel bills you the real token cost and you absorb the gap.
  • 04Caps don't exist by default. AI Usage Limits ship set to keep running and notify. Switch to Block AI at the limit, up to $1,000/month per sub-account.
  • 05Clients can't open the AI Usage Dashboard. You pull their numbers from it, or use SaaS Usage Limits (Labs beta) so they see usage on their own billing page.

The four ways to charge a client for AI

Every agency ends up in one of these, and you can mix them across clients. HighLevel lets you set the rebilling model per product and per sub-account, so one client can be on a fixed rate for chat and a markup for calls.

ModelHow the client is chargedWho carries a usage spikeFits when
Flat fee, AI includedYour monthly fee. AI usage stays on your agency wallet.You, all of itUsage is low and predictable, and you've set a blocking limit
Pass-through at costRebilling with no markup (1x). The client's card refills their own wallet.The clientYou'd rather earn on the service fee and keep usage transparent
MarkupHighLevel's token cost times your multiplier, for example 1.5x or 2xThe client, and their bill moves with every conversationClients who understand usage billing
Fixed rate per unitA flat price per Conversation AI message or Voice AI minute that you setYou, per unit. If the real cost goes above your rate, you pay the difference.Small businesses that want one number they can read
AI Employee plan, resoldGrowth ($50) or Unlimited ($97) per sub-account per month, resold at a price you setNobody until the allowance runs out. Then it pauses, or bills pay per use if Extra Usage is on.Busy agents where the plan costs less than pay per use

Rebilling models from HighLevel's Fixed-Rate Rebilling and AI Employee docs, read Nov 12, 2026. Who carries the risk is my read of how each one bills.

Which HighLevel plan you need to rebill AI

HighLevel's AI docs are clear on this one: agencies must be on the $497 plan (now called Founder) to rebill AI Employee usage, and markup rebilling is $497 only (help docs, Nov 12, 2026). The pricing page lists "Pass-through billing for usage and add-ons" on the $297 Unlimited plan, and its FAQ mentions premium AI usage as rebillable on Unlimited too. Those two don't fully line up, so if you're on $297, open Manage Client > Rebilling on a test sub-account and check whether the AI products are there before you quote anyone.

Phone and email are a separate story: $297 rebills them at cost and $497 adds your markup. I'm on the $497 plan. The full plan comparison is on Pricing, and the SaaS mode side (plans, signups, the configurator) is on SaaS mode guide.

See the $497 Founder plan →

Conversation AI and Voice AI are billed separately now

Until Jul 30, 2026, both products sat under one legacy Conversation AI billing product. HighLevel split them, so each has its own line in billing usage, its own rebilling settings, and its own pricing model. The changelog says it plainly: settings applied to one product won't automatically apply to the other.

So if you set up a client's rebilling before August, open it again. Agency View > AI Suite > Plans & Limits, find the sub-account, three-dot menu > Configure Rebilling, and check both products. A client you thought was on a 2x markup for everything might have Voice AI on whatever the split left it at.

The split is also what makes per-product pricing possible. Chat and calls cost very different amounts, so pricing them the same way rarely makes sense.

How the fixed per-unit price works, and where it loses money

Fixed rate shipped Aug 7, 2026. You pick Fixed Rate instead of Markup for a product, enter a price per unit, and click Save Markups. You can set it in three places: Manage Client > Rebilling for one sub-account, the Rebilling step of a SaaS plan in the SaaS Configurator, or the AI Usage Dashboard. It applies to future charges only. HighLevel's own examples are $0.50 per Conversation AI message and $0.10 per Voice AI minute.

The catch is in HighLevel's FAQ: your agency is still charged the actual token cost, and if your fixed rate is lower than that, "you absorb the difference." Two things push the real cost around while your rate stays put.

Voice AI: the model. As of Nov 10, 2026, speech-to-speech is $0.10 a minute on Gemini 3.8 Live and $0.20 on GPT Realtime 2. Classic voice is $0.045 a minute for the engine plus $0.015 to $0.17 for the voice, plus tokens. An illustration, not a recommendation: if you used HighLevel's example rate of $0.10 a minute and someone later switched the agent to a $0.20 model, every minute would cost you $0.10 more than you bill. On 300 minutes in a month, that's $30 out of your pocket. Phone charges sit on top of all of it on every plan.

Conversation AI: the length of the conversation. It's billed by tokens, and input tokens include the whole conversation history, the instructions and the knowledge base it pulls in. So reply 15 in a long thread costs more than reply 1, while a per-message rate charges them the same. HighLevel's own example conversation (100,000 input and 25,000 output tokens at $1.25 and $10 per million) costs $0.375 in total. Read your real numbers in the Usage Dashboard before you set a rate, and look at your longest conversations, not the average.

If you want fixed rate's simple bill without guessing, lock the model in the agent so nobody upgrades it quietly, and recheck the AI pricing doc when HighLevel adds models. The list changes often.

AI Employee plan or pay per use, per client

The other way to make a client's AI bill predictable is a plan. Growth ($50 a month per sub-account) includes 100 Voice AI minutes and 1,000 Conversation AI responses, and Unlimited ($97) covers both under fair use (AI Product Pricing, Nov 10, 2026). You resell either at your own price in Agency View > Reselling > Subscriptions > AI Employee, then switch it on per client in Manage Client > Reselling.

Do the break-even per client. On a $0.10 a minute voice model, Unlimited pays for itself at about 970 minutes a month. On a $0.20 model it's about 485. Below that, pay per use with rebilling is cheaper for the client. These are the same numbers as Usage fees, worked from the same rates.

Three things plans don't cover: phone charges (always separate), the AI Agent workflow action and Flow Agents (pay per use on every plan). Managed Agents get 100 runs on Growth and 1,000 on Unlimited. If a client's build leans on the AI Agent action, a plan won't make their bill flat. More on that in AI Agent action.

When a Growth allowance runs out, the AI pauses by default. Extra Usage in AI Suite keeps it running at pay-per-use rates. For a receptionist answering real calls, decide that before it happens, because a paused agent at 9pm is a missed call.

Usage spikes and agent loops: how to cap them

Every AI bill horror story is the same shape: something ran far more than anyone expected and nothing was set to stop it. An agent replying to an auto-responder, a workflow that re-triggers itself, a tag that fires a Managed Agent on every update, a voice call nobody hung up. Rebilling doesn't fix that. It just moves the bill to your client, which is worse for the relationship.

HighLevel's caps are all off by default, so stack them.

ControlWhereWhat it does
AI Usage LimitsAgency View > AI Suite > Plans & LimitsMonthly limit per sub-account (up to $1,000) and per user (up to $500). Default is Keep AI running, just notify. Pick Block AI at the limit to stop it. Resets each calendar month.
Extra Usage toggleAI Suite, per sub-account on an AI Employee planOff means the AI pauses when the plan allowance runs out
SaaS Usage Limits (Labs beta, Oct 30, 2026)Manage Client, or as a step when onboarding a SaaS client with rebillingPer-client limit for Voice AI and Conversation AI. You choose bill the extra usage or stop. $497 agencies.
Conversation AI max messagesThe bot's advanced settings1 to 100 per contact. The bot goes inactive for that contact when it hits the limit.
Voice AI call settingsThe agent > Show advanced settings > Call SettingsCall duration and silence handling. The call ends 15 seconds after the idle reminder if the caller stays silent.
Wallet notificationsAgency View > Billing > Notifications > WalletEmails you the same day a sub-account passes a threshold. Doesn't stop anything.

Spend controls from HighLevel's help docs and changelog, read Nov 12, 2026.

Caps inside the agent itself

Account limits are the backstop. The cheaper fix is building the agent so it can't run away in the first place.

Voice AI: in my receptionist build I point people at call settings, which is where you set "how long the calls can be, how short they can be." I also like selling it as an after-hours agent: "after 8:00 p.m., the agent is going to turn on," so it only works the hours the front desk doesn't. And I keep it on its own line: "there's a human number and then there's a voice receptionist number." All three keep minutes tied to calls the business actually wants answered.

Conversation AI: set max messages so one contact can't trigger 100 replies, and on a booking bot, use the option in the booking action where "you can turn the bot off after it makes a booking." Once the job is done, it stops talking. Setup for both is in Train Conversation AI.

Managed Agents and the AI Agent action: every trigger event is a run, so filter the trigger tightly (one form, one tag, one pipeline stage) and test on 10 contacts before a full list. HighLevel's docs list 30 free test runs a month for Managed Agents. The build side is in Agent Studio.

Block or bill the overage?

This is the actual pricing decision, and it's different per agent.

Block when you're on a flat fee or a fixed plan price, because every unit past the line comes out of your margin. Also block on anything outbound or workflow-driven, where a loop or a bad list can burn through money in an hour.

Bill (keep running, or set SaaS Usage Limits to bill the extra usage) when the AI answers inbound leads and a pause costs the client a customer. A receptionist that goes silent mid-month because of a limit does more damage than the overage. Put that in the agreement up front: what the limit is, what happens at it, and what the extra costs.

Either way, the client should hear about it before the limit, not after. With AI Usage Limits, a client who can't manage their own billing sees a "Request more usage" button when they're blocked, and you approve, upgrade or deny it in AI Suite > Plans and Limits > Notifications > Limit Requests. HighLevel's docs say a sub-account is never billed past its limit. Usage figures can lag up to 5 minutes, so leave a little room.

How to report AI usage to clients

The AI Usage Dashboard (Agency View > AI Suite > Usage Dashboard) is agency-only. A client can't open it from their sub-account. So reporting is on you: pick their sub-account, set the range to Month, and you get total transactions, the amount charged to the wallet, tokens, a usage chart and every transaction. Voice AI transactions show call duration, and sometimes zero tokens, because voice bills by the minute.

Their invoice won't explain it either. With rebilling on, the line shows the AI product name and the total for the period, not whether it's a fixed rate or a markup. For a fixed-rate client that's fine, it's units times your price. For a markup client, send a monthly note with minutes, conversations and the total so the number never surprises them.

SaaS Usage Limits fixes most of this for SaaS clients. Their billing page shows usage against their limit, with color cues as they get close and the date it resets. It's Labs beta for Voice AI and Conversation AI only as of Oct 30, 2026, and HighLevel says phone, email and automations are coming, plus alerts at 50%, 80% and 100%.

Rebilling gotchas before you turn it on

No card, no rebilling. If the sub-account doesn't have a payment method when you attach rebilling, it sits On Hold and the client can have restricted access until they pay. Get the card on file first, or you lock your own client out.

Your wallet still pays first. With rebilling on, HighLevel charges the agency wallet, then your Stripe charges the client's card to refill their wallet. A declined client card doesn't stop HighLevel billing you.

Changes aren't retroactive. Switching a client from markup to fixed rate applies to future usage only.

Agency users get Ask AI and AI Studio free under fair use, but a client's own team using them is pay per use unless they have an AI Employee plan. If you give clients logins, that usage shows up too.

Start the 30-day trial and set up rebilling on a test client →

Set up AI rebilling for one client

  1. 01

    Check your plan

    Rebilling AI usage needs the $497 Founder plan per HighLevel's AI docs. On $297, check Manage Client > Rebilling for the AI products before you promise anything.

  2. 02

    Get the client's card on file

    Sub-account Company Billing. Without it, rebilling goes On Hold and the client can lose access.

  3. 03

    Set Conversation AI and Voice AI separately

    AI Suite > Plans & Limits > three-dot menu > Configure Rebilling. Pick Markup or Fixed Rate for each, enter the number, click Save Markups.

  4. 04

    Set a limit and choose what happens at it

    AI Usage Limits on the sub-account. Block AI at the limit for flat-fee and outbound work. Keep running and notify only where a pause costs the client a customer.

  5. 05

    Cap the agent itself

    Call duration on Voice AI, max messages on Conversation AI, bot off after booking, tight triggers on Managed Agents.

  6. 06

    Check the first week in the Usage Dashboard

    Filter to the client, look at the cost per message or minute against your rate, and adjust before the first invoice.

Start your 30-day free trial →

30 days free instead of HighLevel's usual 14, plus a 1:1 setup call with my team, weekly live calls and my agency resources from day one. See everything you get →

Other ways to start

My bonuses come with every one of these.

Best if you're new and want live help setting up

30-day trial + a free live setup bootcamp

Best if you're not sure which plan you need

Compare all 3 plans first

Best if you'll resell HighLevel as your own software

Go straight to the $497 Founder plan

FAQ

Can you rebill Conversation AI and Voice AI in GoHighLevel?

Yes. Turn on rebilling for the sub-account and set each product on its own, since they've been separate billing products since Jul 30, 2026. HighLevel's AI docs say rebilling AI usage needs the $497 Founder plan (as of Nov 12, 2026).

What's the difference between markup and fixed-rate rebilling?

Markup charges the client HighLevel's token cost times your multiplier, so their bill moves with usage. Fixed rate charges a flat price per Conversation AI message or Voice AI minute that you set. HighLevel still bills you the real cost, so with a fixed rate you absorb any gap.

Can I lose money rebilling AI at a fixed rate?

Yes. If the token cost of a message or minute goes above your fixed rate, HighLevel's docs say you absorb the difference. Longer conversations and pricier voice models are the usual causes, so check real costs in the AI Usage Dashboard before setting the rate.

How do I stop a client's AI usage from spiking?

Set an AI Usage Limit on the sub-account and change it to Block AI at the limit, since the default only notifies. Then cap the agent: call duration on Voice AI, max messages on Conversation AI, and tight triggers on Managed Agents.

Can clients see their own AI usage?

Not in the AI Usage Dashboard, which is agency-only. SaaS clients with SaaS Usage Limits (Labs beta since Oct 30, 2026) see usage against their limit on their billing page. Otherwise you pull the numbers and send them.

Is an AI Employee plan cheaper than rebilling pay per use?

Only at volume. Unlimited is $97 a month per sub-account and breaks even around 970 voice minutes on a $0.10 a minute model (rates as of Nov 10, 2026). Phone charges and the AI Agent workflow action are never covered by a plan.

What happens when a client hits their AI limit?

With Block AI at the limit, the AI pauses until you raise the limit or the month resets. A client who can't manage their own billing can click Request more usage, and you approve, upgrade or deny it in AI Suite.

Sources / last checked Nov 12, 2026

  • My videos. My AI receptionist build for med spas (Jun 25, 2026) and my AI employee build for a dental office (Jun 12, 2026). Quoted only, not embedded, because both have income figures in the title or opening.
  • HighLevel rebilling docs. Fixed-Rate Rebilling for AI Products (155000008422, updated Aug 11, 2026), AI Employee: Access, Rebilling and Reselling (155000003906, updated Nov 12, 2026) and Rebilling, Reselling, and Wallets Explained (155000002095, updated Sep 30, 2026) on help.gohighlevel.com, read Nov 12, 2026.
  • HighLevel AI pricing and limits docs. AI Product Pricing (155000006652, updated Nov 10, 2026), AI Usage Limits (155000007813, updated Sep 22, 2026) and AI Usage Dashboard (155000007742, updated Sep 2, 2026), read Nov 12, 2026.
  • HighLevel Voice AI and Conversation AI docs. Create, Test, and Deploy a Voice AI Agent and Configure Idle Reminder Timer for Voice AI (both updated Oct 1, 2026), read Nov 12, 2026. Conversation AI advanced settings and bot status (updated Jul 28 and Oct 12, 2026).
  • HighLevel changelog. ideas.gohighlevel.com/changelog, read Nov 12, 2026: Introducing separate rebilling configuration for Conversation AI and Voice AI (Jul 30, 2026), Set a Fixed Price for Your AI Products (Aug 7, 2026), SaaS Usage Limits (Oct 30, 2026).
  • HighLevel pricing. gohighlevel.com/pricing, read Nov 12, 2026: Unlimited $297/month with pass-through billing for usage and add-ons, Founder $497/month with SaaS mode, AI Employee Growth $50 and Unlimited $97 per sub-account per month.
Hamza Baig (Hamza Automates)

Author

Hamza Baig / Hamza Automates

Founder, Hexona Systems

I run Hexona Systems on GoHighLevel. HighLevel gave it the SaaSpreneur Platinum award in 2024 and Diamond in 2025 and 2026, and I spoke at their LevelUp event in 2026. Hexona powers 1,000+ agencies, we've served 1,500+ businesses, and I've taught 45,000+ students. Everything on this hub is something I use myself or have shown on camera.

Platinum SaaSpreneur
2024
Diamond SaaSpreneur
2025, 2026
LevelUp speaker
2026
YouTube subscribers, Oct 2026
34,200

Keep reading

Start your 30-day free trial →

30 days free + 1:1 setup call + my bonus / affiliate link