HighLevel Updates: New Features for April 13-17, 2026

HighLevel shipped a dense batch of platform updates during the week of April 13 through April 17, 2026. The releases reach nearly every corner of the product: Conversation AI gained trigger-based knowledge retrieval, workflow agents can now call external tools over MCP, Mercado Pago went live as a native payment provider, and Services v2 booking pages picked up several structural upgrades.

Here is a full breakdown of what shipped, what each release does, and where agencies can put it to work. For everything released this month in one place, see the HighLevel updates roundup for April 2026.

AI Agents Gain Conversation Memory, Smart Triggers, and External Tools

Conversation AI now supports Smart Knowledge Base Triggers. Each agent runs one Always On trigger that keeps core brand and offer knowledge available in every conversation, plus up to three Smart Triggers written as plain-English conditions. When a chat reaches the moment a condition describes, the agent pulls answers from the knowledge bases assigned to that trigger. Existing setups were migrated automatically into a default Always On trigger, so nothing changes until you add conditions. In practice, an agency bot can hold back listings, case studies, or offers until a lead is qualified, then surface the right material without being asked.

The AI Agent action in Workflows received two upgrades of its own. A built-in conversation history tool lets the agent read a contact's messaging history across SMS, email, WhatsApp, Instagram, Facebook Messenger, and TikTok, including message content, direction, timestamps, and an AI-generated summary. It is available by default with no configuration. The action also now supports MCP connections, so an agent can call outside services such as Exa for web search, Tavily for search and page extraction, Browserbase for browser automation, or any custom MCP server. Connections support HTTP Streamable and SSE transports, five authentication options, automatic tool discovery, and per-agent tool selection. The AI Agent action is also now cheaper to run. Together these changes let agencies build flows that research and enrich a new lead in seconds instead of minutes of manual lookup. You can explore the full agent lineup on the GoHighLevel AI overview page.

Ask AI Delivers Structured Output and Hands-On Management

Ask AI responses moved beyond plain text. The assistant now renders artifacts such as data tables, charts, code blocks, to-do lists, approvals, media, and web search cards, and it draws Mermaid syntax as visual diagrams directly in chat, complete with full-screen viewing, pan, and zoom. Agencies get pipeline audits, lead source comparisons, and automation flowcharts in a format they can act on rather than reformat.

Ask AI also picked up operational skills. Agencies can create and update SaaS plans, including pricing and configuration, entirely through chat instead of the SaaS Configurator interface. WhatsApp management works the same way: build, edit, or delete Marketing and Utility templates with automatic variable mapping and Meta-required sample values, run single sends and tagged-audience campaigns, set up appointment booking Flows, and pull analytics, all conversationally. If WhatsApp is central to your client work, the GoHighLevel WhatsApp integration now takes far less clicking to manage.

Booking Pages Get Category-First Flows and Conditional Logic

Services v2 booking pages can now open with category selection before any service appears. Categories support images, descriptions, and a card layout, customers only see services inside the category they choose, and empty categories stay hidden. The flow is optional and existing booking pages are unchanged unless you enable it. A separate global setting controls whether service images display at all; turning them off switches the page to a full-width, text-first layout. Another new toggle decides whether customers arriving through a direct service or category link can browse the complete service list, which opens up cross-selling from any entry point.

Custom forms used in calendar booking flows now respect form conditional logic. Disqualification rules run before scheduling, so no appointment is created and no payment is collected for a disqualified lead, and a new Use custom form rules setting controls the post-booking confirmation experience. Support for Services v2 and Rentals is planned for an upcoming update. For salons, spas, clinics, and other service businesses, these changes make bookings easier to browse and keep unqualified appointments off the calendar.

On mobile, a unified appointment screen brings details, attendees, form submissions, notes, activity history, and payment collection into one tabbed view, so staff can charge a card or add notes without leaving the appointment. Granular calendar permission scopes configured on web are now enforced in the mobile app as well, keeping read-only users read-only on every device.

Payments: Mercado Pago Goes Live and Courses Add Tax Support

Mercado Pago is now a native payment provider supporting Colombia, Argentina, Chile, Mexico, Uruguay, Peru, and Brazil, available for both agencies and sub-accounts. Setup uses a Public Key and Access Token plus a webhook configured in the Mercado Pago account, and the Configure Provider feature controls where it is enabled. Minimum transaction amounts apply, and CVV-less payments must be turned on in the Mercado Pago account. Agencies serving Latin American clients can now run checkout on a provider local buyers already trust, with dynamic currency conversion planned next.

Course offers now calculate taxes at checkout based on the buyer's billing address, display a clear breakdown before payment, support tax-inclusive and tax-exclusive pricing, and keep working with coupon codes. Configuration runs through nexus addresses and tax rules in Payments settings. Affiliate Manager also expanded with Pay Per Sale commissions for forms, surveys, and calendars, paid as a flat amount or a percentage when a referred contact completes a purchase within the tracking window. Pay Per Lead and Pay Per Sale stay mutually exclusive per campaign, and calendar tracking requires payments to be enabled.

Funnel AI, AI Studio, and WordPress Builders Level Up

Funnel AI was rebuilt around a multi-stage generation pipeline that plans strategy, page design, and individual sections before generating sections in parallel. Section-specific skill packs cover heroes, navigation, pricing, FAQs, testimonials, stats, and more, and a rebuilt image pipeline runs across GPT and Gemini backends with split, cutout, collage, and overlapping compositions. Cloning now carries customization requests through generation, so clone-with-changes reflects what you asked for. The payoff for agencies is a stronger first draft and fewer manual fixes before a client page goes live.

AI Studio sites with a custom domain can enable pre-rendering, which serves fully rendered HTML to search engines, social platforms, and AI crawlers instead of an empty single-page-application shell. Up to 50 routes per site are cached, served from the edge in under 10 milliseconds, and regenerated on every publish, while regular visitors still get the standard interactive site. That means client sites become indexable on first crawl, show rich social link previews, and can be cited by AI search tools. AI Studio also gained persistent per-project context across builder sessions, history compaction, and faster cloning. On the WordPress side, AI Pages added live color palette previews, per-image regeneration with custom prompts up to 500 characters, and a wider set of section layout variants.

Agency Dashboards for AI Spend, Reselling, and Wallet Activity

A new AI Usage Dashboard gives agency admins one view of AI spend across Conversation AI, Voice AI, Reviews AI, Agent Studio, and Content AI. It reports plan status, total tokens, transaction counts, and pay-per-use charges, with breakdowns by product, model, and billing mode, a detailed transactions table, and sub-account level filtering. Agent Logs gained a Metrics tab with a customizable widget dashboard covering conversations handled, response times, top agents, busiest hours, and node-level action latency, plus auto-refresh intervals from 30 seconds to 15 minutes and filters spanning up to three months.

On the revenue side, the Reselling tab now includes a Location-Level Breakdown showing revenue, cost, profit, and margin per sub-account, with drilldowns into product-level detail and the ability to adjust reselling configurations in place. Agency Wallet added a Summary Tab with total spend, month-over-month change, top spender, and biggest spike cards, a searchable location table, and a product breakdown drawer. Agencies can finally spot loss-making sub-accounts and fix rebilling problems from a single screen.

Workflows Add an Apify Integration and Drag-and-Drop Nodes

Workflows gained native Apify triggers and actions. Triggers fire when an actor or task run finishes, and actions can run tasks or actors, scrape a single URL, fetch dataset items, and retrieve the latest run details. Connection uses an Apify API key, and usage is billed per Apify pricing plus premium workflow charges where applicable. A typical agency use: scrape local businesses from Google Maps and enroll them straight into outreach campaigns without webhooks or third-party connectors.

The workflow canvas itself is easier to reorganize: hovering over a node reveals a six-dot handle you can drag, with a Move here indicator showing valid drop positions, so large workflows no longer require moving steps one at a time through menus.

Smaller Improvements Worth Noting

• Snapshots: the asset selector now shows a nested folder hierarchy with expandable levels, and selecting a folder selects everything inside it, reducing missed assets during pushes.
• Conversations: ID-based search now works across Conversations, Snippets, and Trigger Links, and a WhatsApp number picker on web and mobile lets you choose the sending number when multiple are connected, with 24-hour messaging window rules applied automatically. See our HighLevel WhatsApp guide for setup details.
• Listings: Google Business Profile holiday hours can be managed inside the GBP Optimization tool, with Closed or Split Open days, multiple time intervals, and automatic sync to Google.
• Communities: newsletter send times are now configurable by day, time, and timezone on daily, weekly, bi-weekly, or monthly schedules, and live sessions gained personal meeting links, a recordings hub, and improved noise cancellation.
• Calendars: saving a calendar now scrolls directly to missing required fields, the rentals end-date picker opens in the same month as the start date, assigned user custom values work in the next funnel step and confirmation messages, and round robin distribution is more even.
• Ecommerce: store elements rank at the top of the builder panel when a sub-account has a store.
• Voice AI: the Prompt Optimizer for Voice AI agents is now live, and user create and update APIs now support twilioPhone.
• Labs: four features entered early testing: timezone-aware countdown timers, opportunity forecasting, funnel pages inside courses, and live sessions in courses. Track prerelease features in our HighLevel beta watch.

Frequently Asked Questions

Which countries does the Mercado Pago integration support?

At launch, Mercado Pago works for businesses in Colombia, Argentina, Chile, Mexico, Uruguay, Peru, and Brazil, at both the agency and sub-account level. More regions and dynamic currency conversion are planned, and payments are currently limited to local and regional currencies.

Do Smart Knowledge Base Triggers change my existing Conversation AI setup?

No. Knowledge bases already linked to an agent were moved into a default Always On trigger, and agents behave exactly as before. Smart Triggers only take effect when you add conditions, and each agent supports up to three of them alongside the Always On trigger.

What does an AI Studio site need for pre-render SEO?

Pre-rendering requires a connected custom domain and is enabled per project in settings. After publishing, up to 50 routes are rendered and cached automatically, and the cache refreshes on every publish or domain change. Disabling the feature clears the cache immediately and bots receive the standard single-page application again.

Try These Features Yourself

Every release covered here is available inside the platform now, and the fastest way to evaluate them is hands-on with your own booking flows, workflows, and AI agents. Start a free 30-day GoHighLevel trial and test the features against a real client use case this week.