HubSpot's AI, branded Breeze, is not a flat monthly add-on. It runs on HubSpot Credits, and as of April 14, 2026 the agents most teams deploy are priced on outcomes, so you pay when the work is done rather than for activity along the way. The catch is that "how much does it cost" has two answers: what the agents charge, and what you were already paying for the seats that unlock them. This guide covers both, with every figure verified against HubSpot's own documentation.
The one-line answer
Breeze AI is paid for with HubSpot Credits. Your plan includes a monthly credit allowance, built-in AI features use no credits at all, and the AI agents draw credits down only when they complete a task. Past your allowance you add a credit pack or pay overage at $0.010 per credit. The flagship agents are billed on outcomes: $0.50 per resolved conversation, $1.00 per qualified lead, and about $0.10 per Data Agent answer.
How the credit model works
Credits are the shared currency for HubSpot's usage-based AI. Customer Agent, Prospecting Agent, Data Agent, and Data Studio syncs all draw from the same monthly pool. Four mechanics matter when you budget:
- Included credits come with seat-based plans. The exact allowance varies by subscription and is shown in your account; an Enterprise account, for example, includes 5,000 HubSpot Credits per month.
- Credits reset every month and do not roll over. Unused credits expire at the end of the usage period.
- Only paid seats, meaning Core, Sales, Service, Commerce, and Partner seats, can consume credits. Free and view-only users cannot.
- When included credits run out, usage-based features pause until your reset date, unless you add a credit pack or turn on pay-as-you-go overages.
What actually uses credits, and what is free
This is where most cost estimates go wrong. HubSpot draws a hard line between built-in AI that is included with your plan and agentic AI that completes tasks on your behalf. The included features never touch your credit balance. Credits are spent only when an agent finishes a defined job.
| Uses credits (metered) | Included, no credits |
|---|---|
| Customer Agent resolving a conversation | Breeze Assistant / Copilot |
| Prospecting Agent recommending a lead | Company and contact enrichment |
| Data Agent answering a query | AI summaries and insights |
| Breeze actions inside workflows | AI drafting inside the editor |
In practice this means a team can get real value from Breeze Assistant, enrichment, and summaries without spending a single credit, and only start metering once they turn an agent loose on conversations, leads, or data.
The April 2026 shift: outcome-based agent pricing
On April 14, 2026 HubSpot moved its two flagship Breeze agents to outcome-based pricing. Instead of paying for activity, you pay when the agent completes the job.
Both agents run on Pro and Enterprise, and both now include a 28-day free trial. Because billing is applied in credits, at the $0.010 per credit rate a resolved conversation works out to roughly 50 credits, a recommended lead to roughly 100 credits, and a Data Agent answer to roughly 10 credits.
HubSpot can price this way because Breeze agents sit inside the Smart CRM with your customer data and history, which makes their results consistent enough to guarantee. For context, HubSpot reports Customer Agent already resolves 65% of conversations and cuts resolution time by 39% across more than 8,000 customers who have activated it.
The rate card at a glance
| Breeze feature | You pay for | Price | In credits |
|---|---|---|---|
| Customer Agent | Each resolved conversation | $0.50 | ~50 credits |
| Prospecting Agent | Each lead recommended for outreach | $1.00 | ~100 credits |
| Data Agent | Each answer to a query | ~$0.10 | ~10 credits |
| Credit overage | Each credit beyond your allowance | $0.010 | $10 / 1,000 |
| Built-in AI (Breeze Assistant, enrichment, summaries) | Included in your plan | No credits | Free |
Customer Agent and Prospecting Agent rates are from HubSpot's April 2026 announcement; Data Agent is credit-metered per answer at the commonly cited rate. Built-in AI stays free regardless of volume.
Seats and credits: the other half of the bill
Credits get all the attention, but you cannot spend a credit without a paid seat, so seats are the real base cost. On HubSpot's seat-based model, Sales Hub is priced per seat with no seat minimums:
- Starter: $20 per seat per month
- Professional: $100 per seat per month, plus a one-time $1,500 onboarding fee
- Enterprise: $150 per seat per month, plus a one-time $3,500 onboarding fee
Core seats give edit access and unlock the cross-platform AI; the higher Sales and Service seats add role-specific features. The practical point for budgeting: your monthly HubSpot AI cost is seats plus credits, and for a small team the seats are usually the larger line until agent volume scales.
What it actually costs, in scenarios
The rate card is the easy part. Your bill is a function of volume, so it helps to see the model at real numbers.
| Scenario | Monthly volume | Credits used | Illustrative cost* |
|---|---|---|---|
| Support desk, Customer Agent | 1,000 resolved conversations | ~50,000 | ~$500 |
| Small sales team, Prospecting Agent | 300 recommended leads | ~30,000 | ~$300 |
| Data Agent, ongoing enrichment | 2,000 answers | ~20,000 | ~$200 |
| Light usage, inside included credits | ~50 resolved / 20 leads | ~4,500 | Covered, no extra cost |
*Illustrative at the outcome rates above; actual cost depends on your included credits, tier, and how agents are configured. The pattern is the point: a high-volume support desk can run into the hundreds of dollars a month in credits, while a lean team operating inside its included allowance may pay nothing extra.
What this looks like in a real rollout
We do not have to theorize. INSIDEA recently built an agentic revenue engine for Ruli.AI on HubSpot Sales Hub Professional, with the Prospecting Agent live from day one. In that build the Prospecting Agent consumed 100 HubSpot Credits per lead when outreach was recommended, which lines up exactly with the $1.00 per recommended lead rate. Across the rollout the agents enriched 50,000-plus company records and enrolled 700-plus contacts, and we set a credit cap at kickoff so spend never ran ahead of value.
The lesson most rate-card articles miss: your cost is decided by configuration, not the per-unit price. An agent pointed at a clean, well-scoped list is efficient. An agent pointed at a noisy database burns credits recommending leads you would never call. Setup decides the bill.
How to control credit spend
HubSpot gives you the levers. Using them well is the difference between predictable and surprising invoices:
- Set a maximum monthly credit limit so features pause rather than overspend.
- Choose deliberately between a credit pack, which gives a predictable monthly limit, and pay-as-you-go overage at $0.010 per credit, which resets each month. Packs suit steady usage; pay-as-you-go suits spiky, experimental usage.
- Watch the 75, 85, and 90 percent usage alerts that go to billing contacts and super admins.
- Lean on the free, built-in AI first, and meter only the agent tasks that clearly pay back.
- Scope agents to clean, qualified data before you scale enrollment. This is the single biggest cost lever, and it is a decision you make before go-live.
Is it worth it?
For most teams, yes, provided the setup is disciplined. Outcome-based pricing removes the old risk of paying for AI that does not perform, and at $0.50 per resolved conversation or $1.00 per qualified lead the unit economics compare well against the human hours they replace. The risk is not the price. It is deploying agents on messy data or unbounded lists and letting consumption drift. That is a solvable configuration problem, and the place to solve it is before you switch the agents on.
Get your HubSpot AI spend forecast before you switch it on
INSIDEA is the world's #1 rated Elite HubSpot Partner, and we configure Breeze agents for outcome, not just activation. We scope the data, set credit guardrails, forecast monthly consumption against real volume, and stand up the agents so they pay back from day one. If you want the full HubSpot picture first, start with our HubSpot pricing guide.
How INSIDEA works
