43 products Sparkpulse tracks are shipping around tidymodels. The highest-velocity tidymodels products right now are pysparklyr, factoextra and themis (ranked by Sparkpulse's velocity score). The theme runs across 3 sectors, so the momentum you see here is drawn from products competing in different corners of the market. Their velocity scores span 0.0 to 3.8 out of 10, a spread that shows how unevenly shipping cadence is distributed across the field. Below: every tracked product carrying this theme, updated from release data. Everything on this page is regenerated from verified release data, so the ranking reflects the latest changelogs rather than a static list.
#01pysparklyrPosit's Spark Connect bridge keeps adding backends — and now runs tidymodels tuning on the cluster.3.8alternatives →#02factoextrafactoextra woke from six years of silence and stopped being a FactoMineR front-end3.8alternatives →#03themisthemis is back to adding real resampling algorithms after a documentation-heavy stretch.2.5alternatives →#04tabnetA tabular deep-learning model in R that keeps widening what counts as a tabular task.2.5alternatives →#05modeltimemodeltime built conformal intervals in, then went quiet on features.0.0alternatives →#06tidymodelsThe meta-package ships almost nothing, which is exactly what a version-pinning shim should do0.0alternatives →#07waywiserSpatial model assessment that spent the last year on cross-platform arithmetic and CRAN rules.0.0alternatives →#08dialsdials is quietly registering the tuning parameters for tidymodels' deep-learning push0.0alternatives →#09bundleFour releases in three years, each one teaching the serializer about a model type it couldn't carry0.0alternatives →#10rsampletidymodels' resampling package is retiring its old splitters for sliding windows.0.0alternatives →#11offsetregThe parsnip extension for exposure models grew from one algorithm to three0.0alternatives →#12workflowsetsworkflowsets keeps widening what counts as a model worth comparing.0.0alternatives →#13hardhathardhat keeps adding the contracts tidymodels needs next0.0alternatives →#14censoredcensored keeps survival models aligned with parsnip's shifting prediction contracts0.0alternatives →#15bonsaibonsai keeps widening tidymodels' boosted-tree engine bench, catboost most recently0.0alternatives →#16modeltime.ensemblemodeltime.ensemble wakes after four years, and the work is all tune 2.0 compatibility.0.0alternatives →#17modeldataThe tidymodels example-data package grows one dataset at a time, on nobody's schedule0.0alternatives →#18workflowsThe tidymodels pipeline grew a third stage, and it happens after the model runs.0.0alternatives →#19yardstickyardstick made fairness metrics a first-class part of tidymodels evaluation0.0alternatives →#20textrecipesText features finally stay sparse all the way to the model.0.0alternatives →#21cubistThe R port of Quinlan's Cubist gets reproducibility fixes, not new modelling0.0alternatives →#22tidyposteriorA finished Bayesian model-comparison package in pure maintenance mode0.0alternatives →#23finetunefinetune tracks tune's evolving contracts more than it advances racing itself0.0alternatives →#24fastmlfastml added survival modelling and leakage-proof resampling, moving past classification and regression.0.0alternatives →#25poissonregpoissonreg gave its models away to parsnip and kept the glue — now it just keeps glmnet honest.0.0alternatives →#26parsnipparsnip added a whole new regression type, then wired R models to JAX and PyTorch0.0alternatives →#27discrimdiscrim settled into a thin engine shim after handing its model definitions to parsnip.0.0alternatives →#28modeltime.resamplemodeltime.resample exists to keep backtesting working as tidymodels shifts underneath it.0.0alternatives →#29embedembed keeps adding encoding steps while shedding its deep-learning dependencies0.0alternatives →#30tunetune extends tuning past the model itself to postprocessors, and adds a second parallel backend0.0alternatives →#31tidyclusttidyclust just tripled the model types it can fit, and handed finalization back to tune0.0alternatives →#32BORGA cross-validation guard that refuses to run random CV on dependent data unless you insist0.0alternatives →#33healthyR.aiA healthyverse machine-learning helper in maintenance: one new function in three years.0.0alternatives →#34stacksModel stacking in tidymodels, quietly migrating off foreach and onto future0.0alternatives →#35butcherbutcher expands from trimming models to trimming whole tidymodels workflows0.0alternatives →#36recipestidymodels' preprocessing engine learned sparsity, then settled into deprecations.0.0alternatives →#37bruleetidymodels' torch backend grew from MLPs into a tabular deep learning suite with foundation models.0.0alternatives →#38vetiverPosit's MLOps package went quiet for two years, then came back to keep up with recipes.0.0alternatives →#39finntsMicrosoft's automated forecasting framework, still mostly a one-maintainer effort.0.0alternatives →#40probablyThe package that made calibration a step instead of an afterthought.0.0alternatives →
Frequently asked questions about tidymodels
Which SaaS tools ship tidymodels in 2026?
pysparklyr, factoextra, themis, tabnet, modeltime, and 35 more — the tidymodels products Sparkpulse tracks, ranked by shipping velocity from verified changelogs.
Which tidymodels product has the highest shipping velocity?
pysparklyr, with the top velocity score (3.8/10) — Sparkpulse's velocity score is derived from verified release data.
How many products are shipping around tidymodels?
Sparkpulse currently tracks 43 products carrying the tidymodels theme, updated continuously from verified release data.