Model Summary Plots with Estimates and Confidence Intervals

  conf_level = 0.95,
  coef_map = NULL,
  coef_omit = NULL,
  coef_rename = NULL,
  vcov = NULL,
  add_rows = NULL,
  facet = FALSE,
  draw = TRUE,
  background = NULL,



a model or (optionally named) list of models


confidence level to use for confidence intervals


character vector. Subset, rename, and reorder coefficients. Coefficients omitted from this vector are omitted from the table. The order of the vector determines the order of the table. coef_map can be a named or an unnamed character vector (see the Examples section below). If coef_map is a named vector, its values define the labels that must appear in the table, and its names identify the original term names stored in the model object: c("hp:mpg"="HPxM/G").


string regular expression. Omits all matching coefficients from the table using grepl(perl=TRUE). This argument uses perl-compatible regular expressions, which allows expressions such as "Int|ABC" which omits coefficients matching either "Int" or "ABC", and "^(?!.*Intercept)"` which omits every term except the intercept.


named character vector or function which returns a named vector. Values of the vector refer to the variable names that will appear in the table. Names refer to the original term names stored in the model object, e.g. c("hp:mpg"="hp X mpg") for an interaction term. If you provide a function to coef_rename, modelsummary will create a named vector for you by deriving the new variable names from the vector of original term names with your function.


robust standard errors and other manual statistics. The vcov argument accepts six types of input (see the 'Details' and 'Examples' sections below):

  • NULL returns the default uncertainty estimates of the model object

  • string, vector, or (named) list of strings. Omitting or specifying vcov = NULL will return the model's default uncertainty estimates, e.g. IID errors for standard models. Alternatively, use the string "iid" (aliases: "classical" or "constant") to present IID errors explicitly. The strings "HC", "HC0", "HC1" (alias: "stata"), "HC2", "HC3" (alias: "robust"), "HC4", "HC4m", "HC5", "HAC", "NeweyWest", "Andrews", "panel-corrected", "outer-product", and "weave" use variance-covariance matrices computed using functions from the sandwich package, or equivalent method. The behavior of those functions can (and sometimes must) be altered by passing arguments to sandwich directly from modelsummary through the ellipsis (...), but it is safer to define your own custom functions as described in the next bullet.

  • function or (named) list of functions which return variance-covariance matrices with row and column names equal to the names of your coefficient estimates (e.g., stats::vcov, sandwich::vcovHC, function(x) vcovPC(x, cluster="country")).

  • formula or (named) list of formulas with the cluster variable(s) on the right-hand side (e.g., ~clusterid).

  • (named) list of length(models) variance-covariance matrices with row and column names equal to the names of your coefficient estimates.

  • a (named) list of length(models) vectors with names equal to the names of your coefficient estimates. See 'Examples' section below. Warning: since this list of vectors can include arbitrary strings or numbers, modelsummary cannot automatically calculate p values. The stars argument may thus use incorrect significance thresholds when vcov is a list of vectors.


a data.frame (or tibble) with the same number of columns as your main table. By default, rows are appended to the bottom of the table. You can define a "position" attribute of integers to set the row positions. See Examples section below.


TRUE or FALSE. When the 'models' argument includes several model objects, TRUE draws terms in separate facets, and FALSE draws terms side-by-side (dodged).


TRUE returns a 'ggplot2' object, FALSE returns the data.frame used to draw the plot.


A list of 'ggplot2' geoms to add to the background of the plot. This is especially useful to display annotations "behind" the 'geom_pointrange' that 'modelplot' draws.


all other arguments are passed through to the extractor and table-making functions. This allows users to pass arguments directly to modelsummary in order to affect the behavior of other functions behind the scenes. Examples include:

  • broom::tidy(exponentiate=TRUE) to exponentiate logistic regression. Please see the modelsummary vignette on the package website for important technical notes on this topic.

  • performance::model_performance(metrics="RMSE") to select goodness-of-fit statistics to extract using the performance package (must have set options(modelsummary_get="easystats") first).


if (FALSE) {


# single model
mod <- lm(hp ~ vs + drat, mtcars)

# omit terms with string matches or regexes
modelplot(mod, coef_omit = 'Interc')

# rename, reorder and subset with 'coef_map'
cm <- c('vs' = 'V-shape engine',
  'drat' = 'Rear axle ratio')
modelplot(mod, coef_map = cm)

# several models
models <- list()
models[['Small model']] <- lm(hp ~ vs, mtcars)
models[['Medium model']] <- lm(hp ~ vs + factor(cyl), mtcars)
models[['Large model']] <- lm(hp ~ vs + drat + factor(cyl), mtcars)

# add_rows: add an empty reference category

mod <- lm(hp ~ factor(cyl), mtcars)

add_rows = data.frame(
  term = "factory(cyl)4",
  model = "Model 1",
  estimate = NA)
attr(add_rows, "position") = 3
modelplot(mod, add_rows = add_rows)

# customize your plots with 'ggplot2' functions

modelplot(models) +
  scale_color_brewer(type = 'qual') +

# pass arguments to 'geom_pointrange' through the ... ellipsis
modelplot(mod, color = 'red', size = 1, fatten = .5)

# add geoms to the background, behind geom_pointrange
b <- list(geom_vline(xintercept = 0, color = 'orange'),
  annotate("rect", alpha = .1,
    xmin = -.5, xmax = .5,
    ymin = -Inf, ymax = Inf),
  geom_point(aes(y = term, x = estimate), alpha = .3,
    size = 10, color = 'red', shape = 'square'))
modelplot(mod, background = b)