HubSpot HubSQL: SQL Access to Your CRM Data (Private Beta)

HubSQL is a new HubSpot-built feature, announced in private beta on September 8, 2026, that lets you run standard SQL, including joins, filters, and aggregations, directly against your HubSpot CRM data through a dedicated HubSQL API. It's built by HubSpot itself, not a third-party connector, and it targets the exact complaint HubSpot users have posted for years: no native way to query the CRM database without paying an outside tool to copy the data somewhere else first. As of this writing it's invite-only, with no published price, tier list, or general-availability date.

What is HubSpot HubSQL?

HubSQL is HubSpot's own answer to a question its users have been asking in the community forum since at least 2019: can I just connect a normal SQL client to my HubSpot data and query it, the way I would any other database? For most of HubSpot's history the answer was no. The standard guidance was to pull data out through the CRM API or pay for a third-party sync tool that copies HubSpot records into an external warehouse, on top of the cost of the HubSpot subscription itself.

HubSQL changes that by putting a SQL query layer in front of HubSpot's own CRM data, described in HubSpot's own community post as a "unified, AI-native data access layer using standard SQL." You write a query with joins, filters, and aggregations, and you get an answer back, without exporting anything to a separate database first.

One distinction worth making up front, because search results blur it: this is not the same category as "HubSpot SQL integration" products like Skyvia, Stacksync, or Datawarehouse.io. Those tools sync HubSpot data out into an Azure, PostgreSQL, or Snowflake database you manage and pay for separately. HubSQL queries HubSpot's own data directly, with no external database in the middle. If you already use one of those sync tools, HubSQL is a potential replacement for the reason you bought it, not another layer on top of it.

How does HubSpot HubSQL work?

HubSpot's own description, posted to its developer community on September 8, 2026, is short: "HubSQL Public API. HubSQL is HubSpot's unified, AI-native data access layer using standard SQL. You can now express precise questions, joins, filters, and aggregations, about your customer data with a single query. Paired with AI, HubSQL makes self-serve analytics a reality." Early practitioners on LinkedIn describe the same shape from the outside: "real joins, real filters, real aggregations, straight through a dedicated HubSQL API," with the point being that it replaces the old requirement of standing up "a data pipeline, a warehouse, and a BI tool" just to ask a cross-object question.

What HubSpot has not published yet, as of this writing, matters more than what it has. There's no public documentation confirming which CRM objects and properties are queryable, whether custom objects and association data are included, whether HubSQL is read-only or can also write back to records, what authentication model it uses (a private app, a Service Key, or something HubSQL-specific), or what rate limits apply. The "AI-native" framing and the timing next to HubSpot's other 2026 AI-agent infrastructure work, including its Agent CLI and the growing Breeze agent lineup, suggests HubSQL is meant to feed structured, queryable data to AI tools as much as to a person typing SQL by hand, but HubSpot has not stated that connection directly.

Access today runs through a request-access link HubSpot has been circulating via its developer relations team and in person at Unbound. There is no self-serve toggle in a live portal to test.

What does HubSQL do for your business?

If it ships the way it's being described, the audience is anyone who currently answers a cross-object CRM question, "which deals from companies in this segment closed last quarter, joined against their support ticket volume," by exporting lists, dragging them into a spreadsheet, or paying for a warehouse-and-BI stack whose only job is to make HubSpot data queryable. HubSQL's pitch is collapsing that path down to one query, run against HubSpot's own data, with no separate database to maintain.

Our take: the 2019 community thread this page cites is not a curiosity. It has nearly 30,000 views and has been bumped by new replies every few months for years, which tells you this was a real, recurring, unmet need, not a niche request. A native answer to it is worth watching closely, and worth requesting access to if your team already builds reports by hand-joining HubSpot exports.

Who it's NOT for right now, honestly:

• Anyone who needs this in production today. It's a two-day-old private beta with no SLA, no GA date, and a real chance the scope or access model changes before it ships broadly.
• Anyone budgeting for it this quarter. There is no published price or credit model to plan against yet.
• Anyone who needs guaranteed row-level permissions or field-level data governance today. HubSpot hasn't documented how HubSQL respects existing HubSpot user permissions, and until it does, treat query access as broader than a single user's normal CRM view unless HubSpot states otherwise.
• Anyone who needs it to write data back to the CRM. Every public description so far is about asking questions and getting answers; nothing confirms insert, update, or delete support.

How Hubjoy is approaching HubSQL

We don't have a client on this beta yet, nobody outside HubSpot's invite list does, so we're not going to pretend we've deployed it. Here's the honest plan for the moment access opens up for a client portal.

1. Treat the first queries as a proof of concept, not a dependency. Nothing built on a private beta with no published terms should sit on a critical path. The first HubSQL queries we run for a client will answer a reporting question we can also answer another way, so nothing breaks if the beta's shape changes before GA.
2. Confirm the permissions model before trusting the output. The first thing we'll test is whether a HubSQL query run under a given user's credentials returns the same records that user can already see in HubSpot, or more. Until that's confirmed, we won't hand query access to anyone below admin.
3. Compare it against Data Studio and the CRM API before recommending it over either. HubSpot already sells a dataset-and-joins product in Data Studio, gated to Data Hub Professional and up. If HubSQL ends up gated the same way, or priced similarly, it may compete with a tool a client is already paying for rather than replace a manual export. We'll price both before making a call.
4. Watch for the pattern HubSpot has run before. Data Studio launched in beta with its credit model and tier gate settled only after real usage came in, and HubSpot has changed beta pricing and packaging with limited notice on other features. We'll assume HubSQL's eventual terms can differ from anything described at launch, and we'll re-verify before quoting a client anything as fact.

This page is on our refresh list. The moment HubSpot publishes pricing, a tier list, a permissions model, or a GA date, we'll update it and pull in the real practitioner gotchas that only show up once more than a handful of accounts have access.

Frequently asked questions

What is HubSpot HubSQL?

HubSQL is a HubSpot-built feature, in private beta since September 8, 2026, that lets you run standard SQL, including joins, filters, and aggregations, directly against your HubSpot CRM data through a dedicated API, without exporting the data to an external database first.

Is HubSpot HubSQL available yet?

Only as a private beta, invite or request-access based. There is no self-serve way to turn it on in a live HubSpot portal today, no published general-availability date, and no confirmed list of which subscription tiers will get it when it does ship more broadly.

How do I get access to the HubSQL private beta?

HubSpot has been circulating a request-access link through its developer relations team and demoing HubSQL in person at Unbound, its September 2026 conference. There's no public, permanent sign-up page for it in HubSpot's own product or developer documentation yet, so the fastest path is watching HubSpot's developer channels and the community announcements board for the official beta signup.

What HubSpot data can I query with HubSQL?

Not yet documented publicly. HubSpot's own description says it queries "your customer data" with joins, filters, and aggregations, but no source confirms which specific CRM objects, custom objects, associations, or activity data are in scope. Until HubSpot publishes that detail, assume coverage may be partial at this stage of the beta.

Is HubSQL read-only, or can it write data back to the CRM?

Every public description so far frames HubSQL as asking questions and getting answers back, which points to a query, or read, capability. Nothing published confirms or rules out insert, update, or delete support, so treat write access as unconfirmed rather than assumed either way.

Is HubSQL the same as the third-party "HubSpot SQL" integrations I've seen?

No, and this is worth getting right before you shop. Products like Skyvia, Stacksync, or Datawarehouse.io sync a copy of your HubSpot data into a separate database you pay for and manage, such as an Azure SQL Server or Snowflake instance. HubSQL is built by HubSpot and queries HubSpot's own CRM data directly, with no external database required.

Does HubSQL replace HubSpot Data Studio or the Custom Report Builder?

Not confirmed either way. Data Studio already builds joined, cross-source datasets on Data Hub Professional and Enterprise, and the Custom Report Builder is HubSpot's charting layer. HubSQL may end up complementary to both, feeding SQL-literate teams and AI tools directly, or it may overlap with Data Studio's joins-and-filters use case. We'll update this answer once HubSpot positions HubSQL against its own existing tools.

Does HubSQL cost anything, or use HubSpot Credits?

Not announced. HubSpot hasn't published a price, a HubSpot Credit consumption rate, or an included allotment for HubSQL. Given that other recent AI and data features, including Data Studio and the Breeze agents, run on HubSpot's shared credit system, a credit-based model at general availability is plausible but unconfirmed.

Trying to get more out of your HubSpot data right now?

You don't need to wait on a private beta to fix how your portal reports on itself. We'll look at what you're actually trying to answer and show you what's possible with what you have today.

Book your free HubSpot Review