Softwr

Machine Learning · head to head

BigQuery ML vs Cortex

BigQuery ML logo

BigQuery ML

Machine Learning

Machine learning in BigQuery using SQL

From
Free
Free plan
Yes
Cortex logo

Cortex

Developer Tools

Internal developer portal that scores service ownership and maturity rather than just cataloguing services

From
On request
Free plan
No

The short version

  • BigQuery ML starts at $5/mo; Cortex prices on request.
  • Only BigQuery ML lists SQL-based ML; only Cortex lists service catalogue.
  • Where BigQuery ML falls short: not available in BigQuery's Standard edition, so the cheapest tier cannot use it.
  • Prices and features on this page were last checked on 20 November 2026.

How do BigQuery ML and Cortex differ?

BigQuery ML and Cortex differ on starting price (Free against On request), pricing model (Usage-based against Quote) and free tier (Yes against No). Only the attributes on which BigQuery ML and Cortex actually diverge.

Attributes where BigQuery ML and Cortex differ
AttributeBigQuery MLCortex
Starting priceFreeOn request
Pricing modelUsage-basedQuote
Free tierYesNo
PlatformsWebWeb, API
CategoryMachine LearningDeveloper Tools

Not yet recorded: founded for Cortex.

Which features does each one have?

They share no listed features; 10 appear only for BigQuery ML and 8 only for Cortex. Drawn from each product's published feature list. An absence here means we hold no record of it - not that the product lacks it.

Only in BigQuery ML

  • SQL-based ML
  • AutoML Tables
  • Model export
  • Prediction functions
  • Feature preprocessing
  • BigQuery
  • Vertex AI
  • TensorFlow

Only in Cortex

  • Service catalogue
  • Scorecards
  • Initiatives
  • Self-service templates
  • Query language
  • Eng intelligence
  • Integrations layer
  • On-call and ownership mapping

What is each one used for?

BigQuery ML is most often brought in for training models in SQL without exporting data. The jobs each tool is most often brought in to do.

BigQuery ML

  • Training models in SQL without exporting datanot Cortex
  • Linear and logistic regression on warehouse datanot Cortex
  • K-means clustering and matrix factorisation for recommendationsnot Cortex
  • Time series forecasting with ARIMA_PLUSnot Cortex
  • Running imported ONNX, TensorFlow or XGBoost models against BigQuery datanot Cortex

Cortex

  • A platform team that needs to prove to leadership how far a Kubernetes or framework migration has actually progressed across hundreds of servicesnot BigQuery ML
  • An organisation where ownership of production services is genuinely unknown after reorganisations or acquisitionsnot BigQuery ML
  • Enforcing production readiness standards on new services at creation time rather than at incident reviewnot BigQuery ML
  • Replacing a self-hosted Backstage that is consuming a full-time engineer or two just to stay upgradednot BigQuery ML

Where do BigQuery ML and Cortex fall short?

BigQuery ML's most cited limit: not available in BigQuery's Standard edition, so the cheapest tier cannot use it. Documented limitations, not opinions. Every one is a constraint you would hit in normal use.

BigQuery ML

  • Not available in BigQuery's Standard edition, so the cheapest tier cannot use it
  • Billed through BigQuery compute and storage rather than as its own product, so training cost tracks data scanned
  • Remote models incur extra Agent Platform charges on top
  • Externally trained model types such as boosted trees and AutoML run through Agent Platform rather than inside BigQuery

Cortex

  • Pricing is quote only, so a platform team cannot build a business case or compare against the true cost of self-hosted Backstage without entering a sales cycle first
  • The catalogue is only as accurate as the metadata you supply, so an organisation with inconsistent repository conventions spends its first quarter cleaning data before any scorecard means anything
  • Scorecards create political friction: publishing a per-team grade turns an engineering standard into a performance metric, and teams will game the rules rather than fix the underlying issue
  • It is closed source and hosted, so organisations with strict data residency or air-gap requirements have limited options compared with running Backstage themselves
  • Value scales with organisation size, so a company with fifty engineers and thirty services will find the catalogue tells them nothing they did not already know while still paying an enterprise contract

How much do BigQuery ML and Cortex cost?

BigQuery ML has a free plan, with paid plans from $5/mo; Cortex is priced on request. Plan by plan below, from each vendor's own pricing page.

BigQuery ML

Free
  • Free TierFree
    • 10GB storage
    • 1TB queries
  • On-Demand$5/tb
    • Pay per TB scanned
    • ML training costs

Cortex

On request
  • CortexOn request
    • Service catalogue and ownership mapping
    • Scorecards and initiatives
    • Self-service templates

Which should you pick?

Choose BigQuery ML if

  • You need sql-based ml.
  • You want to start without paying.
  • You also want automl tables.

Choose Cortex if

  • You need service catalogue.
  • You work on Web, API.
  • You also want scorecards.

Questions people ask

Is BigQuery ML or Cortex better?
Neither clearly leads. BigQuery ML starts at Free and Cortex at On request, and user ratings are close enough to be indistinguishable. Choose on capability and platform support.
Which is cheaper, BigQuery ML or Cortex?
BigQuery ML has a free tier; the other does not. Paid plans start at Free for BigQuery ML and On request for Cortex.
Does BigQuery ML or Cortex run on more platforms?
BigQuery ML runs on Web. Cortex runs on Web, API.
Can I use BigQuery ML for free?
Yes. BigQuery ML has a free tier, so you can try it without paying. Cortex starts at On request.
What is BigQuery ML best used for?
BigQuery ML is most often used for training models in sql without exporting data, linear and logistic regression on warehouse data, k-means clustering and matrix factorisation for recommendations, time series forecasting with arima_plus. Of those, training models in sql without exporting data and linear and logistic regression on warehouse data are not what Cortex is typically brought in for.
What can BigQuery ML do that Cortex cannot?
BigQuery ML covers SQL-based ML, AutoML Tables, Model export, Prediction functions. Cortex covers Service catalogue, Scorecards, Initiatives, Self-service templates.

Answered from the vendors’ own pages

BigQuery ML: How much does Google Cloud BigQuery ML cost?

BigQuery ML pricing is not specified separately on Google Cloud's pricing page. It follows the same pay-as-you-go model as BigQuery, charging per terabyte of data scanned during analysis. Customers receive $300 in free credits and can use 20+ products free up to monthly limits.

Source
Cortex: How is Cortex different from Backstage?

Backstage is an open source framework you build and staff. Cortex is a hosted product with scorecards and initiatives out of the box and no platform team required to keep it running.

BigQuery ML: Does Google Cloud offer a free trial?

Yes, new customers get $300 in free credits and all customers can use 20+ Google Cloud products free up to their monthly usage limits.

Source
Cortex: Is pricing published?

No. Cortex requires a sales conversation and prepares a custom proposal.

Cortex: Can it self-host?

It is offered as a hosted product; on-premise arrangements are handled through sales rather than published.

Cortex: What is the main adoption risk?

Bad or missing service metadata. The catalogue and every scorecard on top of it inherit whatever quality your repositories already have.

Share

Related pages

Other head to heads