diff --git a/NAMESPACE b/NAMESPACE
index e5df44c7..138ecf21 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -51,6 +51,8 @@ export(activation)
 export(activation_2)
 export(adjust_deg_free)
 export(all_neighbors)
+export(average_before_softmax)
+export(balance_probabilities)
 export(batch_size)
 export(buffer)
 export(cal_method_class)
@@ -119,6 +121,7 @@ export(no_global_pruning)
 export(num_breaks)
 export(num_clusters)
 export(num_comp)
+export(num_estimators)
 export(num_hash)
 export(num_knots)
 export(num_leaves)
@@ -172,6 +175,7 @@ export(shrinkage_variance)
 export(signed_hash)
 export(significance_threshold)
 export(smoothness)
+export(softmax_temperature)
 export(spline_degree)
 export(splitting_rule)
 export(stop_iter)
@@ -182,6 +186,7 @@ export(svm_margin)
 export(target_weight)
 export(threshold)
 export(token)
+export(training_set_limit)
 export(tree_depth)
 export(trees)
 export(trim_amount)
diff --git a/NEWS.md b/NEWS.md
index c1384934..b9e49d5c 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -2,6 +2,8 @@
 
 * A bug was fixed where some space-filling designs did not respect the `original` argument (#409).
 
+* Parameters were added for the `tab_pfn` model: `num_estimators()`, `softmax_temperature()`, `balance_probabilities()`, `average_before_softmax()`, and `training_set_limit()`.
+
 # dials 1.4.2
 
 * `prop_terms()` is a new parameter object used for recipes that do supervised feature selection (#395). 
diff --git a/R/param_schedulers.R b/R/param_schedulers.R
index 1cfb5fcc..f0824f05 100644
--- a/R/param_schedulers.R
+++ b/R/param_schedulers.R
@@ -1,5 +1,5 @@
 #' Parameters for neural network learning rate schedulers
-#
+#'
 #' These parameters are used for constructing neural network models.
 #'
 #' @inheritParams Laplace
diff --git a/R/param_tab_pfn.R b/R/param_tab_pfn.R
new file mode 100644
index 00000000..36fd3d5b
--- /dev/null
+++ b/R/param_tab_pfn.R
@@ -0,0 +1,70 @@
+#' Parameters for TabPFN models
+#'
+#' These parameters are used for constructing Prior data fitted network (TabPFN)
+#' models.
+#'
+#' @inheritParams Laplace
+#' @inheritParams select_features
+#'
+#' @details
+#' These parameters are often used with TabPFN models via `parsnip::tab_pfn()`.
+#' @name tab-pfn-param
+#' @export
+num_estimators <- function(range = c(1, 25), trans = NULL) {
+  new_quant_param(
+    type = "integer",
+    range = range,
+    inclusive = c(TRUE, TRUE),
+    trans = trans,
+    label = c(num_estimators = "# Estimators"),
+    finalize = NULL
+  )
+}
+
+#' @rdname tab-pfn-param
+#' @export
+softmax_temperature <- function(range = c(0, 10), trans = NULL) {
+  new_quant_param(
+    type = "double",
+    range = range,
+    inclusive = c(FALSE, TRUE),
+    trans = trans,
+    label = c(softmax_temperature = "Softmax Temperature"),
+    finalize = NULL
+  )
+}
+
+#' @rdname tab-pfn-param
+#' @export
+balance_probabilities <- function(values = c(TRUE, FALSE)) {
+  new_qual_param(
+    type = "logical",
+    values = values,
+    label = c(balance_probabilities = "Balance Probabilities?"),
+    finalize = NULL
+  )
+}
+
+#' @rdname tab-pfn-param
+#' @export
+average_before_softmax <- function(values = c(TRUE, FALSE)) {
+  new_qual_param(
+    type = "logical",
+    values = values,
+    label = c(average_before_softmax = "Average Before Softmax?"),
+    finalize = NULL
+  )
+}
+
+#' @rdname tab-pfn-param
+#' @export
+training_set_limit <- function(range = c(2L, 10000L), trans = NULL) {
+  new_quant_param(
+    type = "integer",
+    range = range,
+    inclusive = c(TRUE, TRUE),
+    trans = trans,
+    label = c(training_set_limit = "Training Set Size"),
+    finalize = NULL
+  )
+}
diff --git a/_pkgdown.yml b/_pkgdown.yml
index 5a19beef..d1ec29a2 100644
--- a/_pkgdown.yml
+++ b/_pkgdown.yml
@@ -82,6 +82,7 @@ reference:
   - neighbors
   - num_clusters
   - num_comp
+  - num_estimators
   - num_knots
   - penalty
   - predictor_prop
diff --git a/man/scheduler-param.Rd b/man/scheduler-param.Rd
index 0410054c..867c54bf 100644
--- a/man/scheduler-param.Rd
+++ b/man/scheduler-param.Rd
@@ -11,8 +11,7 @@
 \alias{rate_decay}
 \alias{rate_schedule}
 \alias{values_scheduler}
-\title{Parameters for neural network learning rate schedulers
-These parameters are used for constructing neural network models.}
+\title{Parameters for neural network learning rate schedulers}
 \format{
 An object of class \code{character} of length 5.
 }
@@ -47,7 +46,6 @@ transformation, \code{NULL}.}
 in examples below.}
 }
 \description{
-Parameters for neural network learning rate schedulers
 These parameters are used for constructing neural network models.
 }
 \details{
diff --git a/man/tab-pfn-param.Rd b/man/tab-pfn-param.Rd
new file mode 100644
index 00000000..ead8a341
--- /dev/null
+++ b/man/tab-pfn-param.Rd
@@ -0,0 +1,40 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/param_tab_pfn.R
+\name{tab-pfn-param}
+\alias{tab-pfn-param}
+\alias{num_estimators}
+\alias{softmax_temperature}
+\alias{balance_probabilities}
+\alias{average_before_softmax}
+\alias{training_set_limit}
+\title{Parameters for TabPFN models}
+\usage{
+num_estimators(range = c(1, 25), trans = NULL)
+
+softmax_temperature(range = c(0, 10), trans = NULL)
+
+balance_probabilities(values = c(TRUE, FALSE))
+
+average_before_softmax(values = c(TRUE, FALSE))
+
+training_set_limit(range = c(2L, 10000L), trans = NULL)
+}
+\arguments{
+\item{range}{A two-element vector holding the \emph{defaults} for the smallest and
+largest possible values, respectively. If a transformation is specified,
+these values should be in the \emph{transformed units}.}
+
+\item{trans}{A \code{trans} object from the \code{scales} package, such as
+\code{scales::transform_log10()} or \code{scales::transform_reciprocal()}. If not provided,
+the default is used which matches the units used in \code{range}. If no
+transformation, \code{NULL}.}
+
+\item{values}{A vector of possible values (TRUE or FALSE).}
+}
+\description{
+These parameters are used for constructing Prior data fitted network (TabPFN)
+models.
+}
+\details{
+These parameters are often used with TabPFN models via \code{parsnip::tab_pfn()}.
+}
diff --git a/tests/testthat/test-params.R b/tests/testthat/test-params.R
index 431faf52..dd110e5b 100644
--- a/tests/testthat/test-params.R
+++ b/tests/testthat/test-params.R
@@ -143,6 +143,15 @@ test_that("param ranges", {
   expect_equal(mtry_prop(c(.1, .2))$range, list(lower = .1, upper = .2))
   expect_equal(dropout(c(.1, .2))$range, list(lower = .1, upper = .2))
   expect_equal(prop_terms(c(.1, .2))$range, list(lower = .1, upper = .2))
+  expect_equal(num_estimators(c(1L, 10L))$range, list(lower = 1L, upper = 10L))
+  expect_equal(
+    softmax_temperature(c(0.1, 2.0))$range,
+    list(lower = 0.1, upper = 2.0)
+  )
+  expect_equal(
+    training_set_limit(c(2L, 10L))$range,
+    list(lower = 2L, upper = 10L)
+  )
 })
 
 
@@ -179,4 +188,6 @@ test_that("param values", {
   expect_equal(all_neighbors(TRUE)$values, TRUE)
   expect_equal(cal_method_class()$values, values_cal_cls)
   expect_equal(cal_method_reg()$values, values_cal_reg)
+  expect_equal(balance_probabilities(TRUE)$values, TRUE)
+  expect_equal(average_before_softmax(TRUE)$values, TRUE)
 })