The Complete HubSpot Breeze Agent Guide for GTM Teams

By the time a B2B buyer signals intent, they’ve already done most of the work. They’ve read the comparison pages, watched the demos, and shortlisted their options. What they haven’t done is raise their hand. And the window between signal and outreach, i.e., between a prospect being in-market and a rep actually reaching them, is where most pipelines leak.

Workflows weren’t built to close that window. They were built to nurture contacts who had already identified themselves, through paths you already mapped. Anything outside that map lands in a gap.

Breeze Agents operate in the gap. They monitor signals, enrich records, draft outreach, resolve tickets, and generate content. Not because you triggered them to, but because that’s the goal they’re working toward.

HubSpot now has 20+ of them spanning every GTM function. This guide breaks down the full roster: by hub, by availability status, and by what they actually need from your CRM to deliver on the promise.

What Are HubSpot AI Agents? 

Inside HubSpot’s ecosystem, there are three distinct layers of AI capability.

Breeze AI Assistant responds to prompts. You ask; it answers. It helps summarize records, draft emails, and answer questions about your CRM. Assistants are useful, but reactive by design.

Workflow Automation executes predefined rules. For example, when a contact fills out a form, a sequence triggers. The logic is yours; HubSpot just runs it. However, there’s no adaptation for scenarios you didn’t anticipate when building the workflow.

HubSpot AI Agents are different from both. They receive a goal, decompose it into steps, use tools and data to execute, and adjust when something doesn’t work. They don’t wait to be asked and don’t quit when the situation diverges from a predefined path.

HubSpot Breeze AI is the umbrella layer that powers all agents, assistants, and the 100+ embedded AI features spread across every hub. Breeze Agents sit at the top of that stack with an autonomous and goal-directed mechanism.

What makes HubSpot’s approach distinct is its grounding. Breeze Agents don’t operate in isolation. They pull from the same Smart CRM that your reps, marketers, and service teams work in every day.

Your customer conversations, deal history, web activity, and campaign data serve as accessible context. The result is an enriched output that differentiates HubSpot agents from generic marketplace ones.

HubSpot AI Agents by Hub

HubSpot’s Breeze Marketplace hosts 20+ agents and assistants spanning marketing, sales, service, and operations. They are organized by hub. Each agent lives in the function it serves and is designed to own a specific slice of the GTM workflow.

Marketing Hub Agents

Content Agent (GA) — It generates landing pages, blogs, podcasts, and case studies in your brand voice, which are informed by Smart CRM data. It’s the most-used HubSpot AI agent in the marketing suite, but output quality is directly proportional to how well your Brand Knowledge section is configured.

Social Media Agent (Beta) — It analyzes your social presence and creates posts based on your company details, audience, industry, and marketing best practices. It is useful for teams that need consistent social output without dedicating bandwidth to daily content production. 

Blog Research Agent (Beta) — It surfaces relevant research, data points, and topic angles to support blog production. This reduces the time writers spend on background research before drafting. 

Sales to Marketing Feedback Agent (Beta) — It turns field intelligence from sales conversations into actionable marketing input. In practice, this closes one of the most persistent gaps in B2B GTM. 

ABM Landing Page Agent (Beta) — It builds account-specific landing pages using CRM data, tailored to the company, industry, or persona being targeted. This automates high-intent, high-effort work that previously required manual customization at every account. 

Sales Hub Agents

Prospecting Agent (GA) — The most commercially mature agent in the Breeze suite, it monitors buying signals, identifies decision-makers, researches target accounts, and writes personalized outbound emails on behalf of a connected rep. 

Early users of the Prospecting Agent report outreach response rates 2x the industry benchmark.^[i] 

Company Research Agent (Beta) — It surfaces funding rounds, hiring signals, recent news, and tech stack data for target accounts directly into the CRM record. This cuts the pre-call research that typically eats up hours per week.

Call Recap Agent (Beta) — It generates structured summaries of sales calls, capturing key discussion points, objections, next steps, and CRM updates automatically. This removes the post-call admin that most reps deprioritize.

Closing Agent (Beta) — It assists reps in late-stage deal management. This includes tracking commitments, flagging stalled deals, and surfacing next-best actions to push toward a close.

Deal Loss Agent (Beta) — It analyzes patterns across lost deals and surfaces recommendations to improve win rates. This turns loss data into actionable intelligence.

RFP Agent (Beta) — It auto-fills requests for proposals based on past responses stored in HubSpot. A process that typically consumes days of cross-functional coordination gets compressed significantly.

Cross-sell / Upsell Agent (Beta) — It identifies expansion opportunities within existing accounts by analyzing product usage, engagement signals, and customer health data. This brings a layer of proactive intelligence to a motion that most teams run reactively.

Customer Service / Success Agents

Customer Agent (GA) — It resolves customer inquiries automatically across 9 channels, including email, WhatsApp, SMS, and a beta Voice/Calling channel. It pulls from your knowledge base, CRM tickets, and conversation history to provide on-brand, accurate responses. 

HubSpot reports teams using the Customer Agent alongside Service Hub see 70% increase in ticket resolution rates.^[ii]

Customer Health Agent (Beta) — It assesses customer health scores across accounts and surfaces next-best actions for CS teams. This shifts customer success from reactive firefighting to proactive risk management.

Customer Handoff Agent (Beta) — It manages account transitions between owners using CRM and interaction data, preserving context that typically gets lost in manual handoffs. This is particularly relevant for teams with high rep turnover or frequent territory realignment.

Knowledge Base Agent (Beta) — It works alongside the Customer Agent to identify gaps in your support content. It analyzes tickets and conversations to flag where documentation is failing customers, then drafts Knowledge Base articles to fill those gaps. 

CRM & Operations

Data Agent (GA) — This is the intelligence engine underneath the rest of the suite. It researches and answers custom questions about your customers by analyzing CRM data, call recordings, and documents. Using Smart Properties, it finds new information about prospects from across the web and adds it directly to the Smart CRM. 

Developer

Developer Tool Testing Agent (Beta) — It automates testing workflows for teams building on HubSpot’s developer platform. This reduces manual QA cycles for portal admins and developers managing custom integrations.

Ecommerce

Shopify Store Performance Agent (GA) — It pulls a complete performance view of your connected Shopify store directly inside HubSpot, surfacing key metrics, summarizing trends, and recommending actions without requiring manual analysis or a separate analytics tool.

Is Your HubSpot Portal AI Ready?

Implementing HubSpot AI agents isn’t just about switching on Breeze. Their effectiveness depends on the quality of your CRM, the consistency of your processes, and the trustworthiness of the data they rely on.

Before deploying HubSpot Breeze Agents, evaluate your portal across these five areas:

1. Data Readiness

AI is only as good as the data it can access. Review your CRM for duplicate records, incomplete properties, outdated contacts, and inconsistent naming conventions. 

2. Content Readiness

Several HubSpot AI agents generate content using your existing assets. Ensure your knowledge base, website copy, product documentation, and brand guidelines are up-to-date. 

3. Process Readiness

AI performs best when your business processes are clearly defined. Standardized lifecycle stages, documented sales pipelines, and service workflows provide the structure agents need to automate tasks confidently.

4. Governance & Security

When deploying agents, ensure your team is aware of the current HubSpot data policy, which prioritizes customer privacy and restricts the usage of sensitive customer data for model training purposes. Building on a foundation of explicit consent and clear data governance is now a prerequisite for any agentic CRM strategy.

5. Measurement & Optimization

Don’t stop at deployment. Track adoption, response quality, productivity gains, and business outcomes to understand which HubSpot Breeze Agents deliver the most value. Use these insights to refine prompts, processes, and configurations over time.

The Future of AI in HubSpot

HubSpot’s vision for AI extends beyond individual assistants/agents. It’s building an ecosystem where specialized HubSpot Breeze Agents collaborate to support the entire customer lifecycle.

Multi-Agent Workflows 

The Run Agent workflow action, currently in beta, lets you trigger agents from any CRM event. This is the connective tissue that turns individual agents into coordinated systems. For example: A lead downloads a technical guide; the Data Agent enriches the record; a scoring model flags high intent; and the Prospecting Agent drafts outreach — all within a single automated flow. 

Marketplace Expansion 

With 20+ agents already available and HubSpot opening Breeze Studio to third-party developers, the Marketplace will expand significantly. Expect more industry-specific agents, deeper integrations with external tools via HubSpot Model Context Protocol, and purpose-built agents for niche GTM workflows. 

Agent Collaboration Across Hubs

As more agents reach GA and the orchestration layer matures, handoffs between hub-specific agents will become standard. For example, the Sales to Marketing Feedback Agent can pass field intelligence to the Content Agent, while the Customer Health Agent flags expansion signals for the Cross-sell Agent to act on. 

AI-driven Operations Become the Default 

Breeze Studio now allows teams to build custom agents trained on their own processes. The organizations that do this early will operate at a different level than those still running HubSpot as a workflow tool.  

Human-in-the-loop as standard practice  

The most effective deployments today aren’t fully autonomous. They’re structured so agents handle execution volume while humans retain ownership of strategic judgment. That model is becoming the operational baseline for mature HubSpot implementations.

Statistics References:

[i] HubSpot Spring 2026 Spotlight

[ii] HubSpot Breeze Customer Agent