Text

Ren'Py contains several ways of displaying text. The say and menu statements are primarily concerned with the display of text to the user. The user interface often contains text, displayed using the text, textbutton, and label screen language statements. These functions, along with others, create Text() displayables, and show them on the screen.

The Text displayable is responsible for managing the process of showing the text to the user. The text displayable performs actions in the following order:

Translating text.
Interpolating data into the text.
Styling the text using styles and text tags.
Laying out the styled text.
Drawing the text to the screen.

This chapter discusses the process of text display in Ren'Py.

Escape Characters

There are three special characters that can control the way Ren'Py displays text. A creator needs to be aware of these characters to ensure that their writing is not accidentally misinterpreted by the engine.

(backslash)

The backslash character is used to introduce when writing a Ren'Py or Python string. Some common escape codes are:

\" (backslash-doublequote): Includes a doublequote in a double-quoted string.
\' (backslash-quote): Includes a single quote in a single-quoted string.
\ (backslash-space): Includes an additional space in a Ren'Py string. By default, Ren'Py script text collapses adjacent whitespace into a single space character.
\n (backslash-n): Includes a newline character in the text.
\\ (backslash-backslash): Includes a backslash character in the text.
\% (backslash-percent): Includes a protected percent character in the text. It's also possible to write it as %% : both protections will result in a single % character being written in the end.

[ (left bracket)

The left bracket is used to introduce interpolation of a value into the text. To include a single left bracket in your text, double it – write [[.

{ (left brace)

The left brace is used to introduce a text tag. To include a left brace in your text, double it – write {{.

【 (left lenticular bracket)

The left lenticular bracket is used to to introduce ruby/furigana text. To include a left lenticular bracket in your text, double it – write 【【.

Interpolating Data

Ren'Py supports interpolating data into the text string before it is displayed. For example, if the player's name is stored in the playername variable, one could write a line of dialogue like:

g "Welcome to the Nekomimi Institute, [playername]!"

Ren'Py will search for variables in the following order:

When used in a screen, screen-local variables.
Variables found in the interpolate namespace.
Variables found in the global namespace.

Ren'Py isn't limited to interpolating simple variables. It can also interpolate any valid Python expression. So it's possible to have:

g "My first name is [player.names[0]]."

It's possible to apply formatting when displaying numbers. This will display a floating point number to two decimal places:

g "I like you [100.0 * points / max_points:.2] percent!"

Ren'Py's string formatting is taken from the PEP 3101 string formatting syntax. Ren'Py uses [ to introduce string interpolation because { was taken by text tags.

Along with the !s and !r conversion flags supported by Python, Ren'Py supports several more flags. The !q conversion flag ensures that text tags are properly quoted, so that displaying a string will not introduce unwanted formatting constructs. For example:

g "Don't pull a fast one on me, [playername!q]."

The !t flag will translate the interpolated string:

if points > 5:
    $ mood = _("happy")
else:
    $ mood = _("annoyed")

g "I'm [mood!t] to see you."

The !i flag will make additional interpolate for the interpolated string:

define earned_points_info = _("[points]{image=points.png} earned points")
g "I'm happy to see you you have [earned_points_info!ti]."

This should be used to substitute the text that has a substitution inside. It's often useful in screen language, see Const Text.

The !u flag forces the text to uppercase and the !l flag forces the text to lowercase. The !c flag acts only on the first character, capitalizing it. These flags may be combined, for example using !cl would capitalize the first character, and force the remaining text to lowercase.

It should be noted that:

the order in which the flags are given does not change the result : !cl will do just the same as !lc.
Supplementary exclamation marks will be ignored, and will not circumvent the previous rule : !l!c will do the same as !c!l or !cl.

The transformations are done in the following order:

r/s (repr or str)
t (translate)
i (recursive interpolation)
q (quoting)
u (uppercase)
l (lowercase)
c (capitalize)

Styling and Text Tags

In Ren'Py, text gains style information in two ways. The first is from the style that is applied to the entire block of text. Please see the section about the style system for more details, especially the section on text style properties.

The second way is through text tags. Text tags are suitable for styling a portion of text block, or a small fraction of the text blocks in the program. If you find yourself applying the same text tags to every line of text, consider using a style instead.

There are two types of text tags. Some text tags are self-closing, while others allow a closing tag. The close tag starts with a slash (/), and tags are closed in last-in, first-out order. Each closing tag should match the corresponding opening tag. If not, Ren'Py will produce a lint warning.

For example:

# This line is correct.
"Plain {b}Bold {i}Bold-Italic{/i} Bold{/b} Plain"

# This line is incorrect, and will cause incorrect behavior.
"Plain {b}Bold {i}Bold-Italic{/b} Italic{/i} Plain"

It is not necessary to close all text tags. Ren'Py will close all tags that are open at the end of the text block. For example:

"{size=+20}This is big!"

Some text tags take an argument. In that case, the tag name is followed by an equals sign (=), and the argument. The argument may not contain the right-brace (}) character. The meaning of the argument varies based on the text tag.

General Text Tags

Tags that apply to all text are:

a

The anchor tag creates a hyperlink between itself and its closing tag. While the behavior of the hyperlink is controlled by the hyperlink_functions style property, the default handler has the following behavior.

When the argument begins with jump:, the rest of the argument is a label to jump to.
When the argument begins with call:, the rest of the argument is a label to call. As usual, a call ends the current Ren'Py statement.
When the argument begins with call_in_new_context:, the rest of the argument is a label to call in a new context (using renpy.call_in_new_context()).
When the argument begins with show:, the rest of the argument is a screen to show.
When the argument begins with showmenu:, the rest of the argument is a game menu screen to show.
Otherwise, the argument is a URL that is opened by the system web browser.

If there is no protocol section in the argument, config.hyperlink_protocol is prepended to it. If config.hyperlink_protocol has been set to "jump", {a=label} and {a=jump:label} become equivalent. Creators can define new protocols using config.hyperlink_handlers.

label test:

    e "Why don't you visit {a=https://renpy.org}Ren'Py's home page{/a}?"

    e "Or {a=jump:more_text}here for more info{/a}."

    return

label more_text:

    e "In Hot Springs, Arkansas, there's a statue of Al Capone you can take a picture with."

    e "That's more info, but not the kind you wanted, is it?"

    return

alpha

The alpha text tag renders the text between itself and its closing tag in the specified opacity. The opacity should be a value between 0.0 and 1.0, corresponding to fully invisible and fully opaque, respectively. If the value is prefixed by + or -, the opacity will be changed by that amount instead of completely replaced. If the value is prefixed by *, the opacity will be multiplied by that amount.

"{alpha=0.1}This text is barely readable!{/alpha}"
"{alpha=-0.1}This text is 10 percent more transparent than the default.{/alpha}"
"{alpha=*0.5}This text is half as opaque as the default.{/alpha}"

alt

The alt tag prevents text from being rendered, while still making the text available for the text-to-speech system.

g "Good to see you! {image=heart.png}{alt}heart{/alt}"

Dialogue Text Tags

Text tags that only apply to dialogue are:

w

The wait tag is a self-closing tag that waits for the user to click to continue. If it is given an argument, the argument is interpreted as a number, and the wait automatically ends after that many seconds have passed.

"Line 1{w} Line 1{w=1.0} Line 1"

p

The paragraph pause tag is a self-closing tag that terminates the current paragraph, and waits for the user to click to continue. If it is given an argument, the argument is interpreted as a number, and the wait automatically ends after that many seconds have passed.

"Line 1{p}Line 2{p=1.0}Line 3"

nw

The no-wait tag is a self-closing tag that causes the current line of dialogue to automatically dismiss itself once the end of line has been displayed.

g "Looks like they're{nw}"
show trebuchet
g "Looks like they're{fast} playing with their trebuchet again."

If it is given an argument, the argument is interpreted as a number, and the wait automatically ends after that many seconds have passed.

g "I'm gonna fall in a few seconds!{nw=2}"
show g_gone

The no-wait tag will wait for self-voicing to complete before advancing.

fast

If the fast tag is displayed in a line of text, then all text before it is displayed instantly, even in slow text mode. The fast tag is a self-closing tag.

g "Looks like they're{nw}"
show trebuchet
g "Looks like they're{fast} playing with their trebuchet again."

If present, the {fast} tag will cause voice to be sustained from the previous line.

done

Text after the done tag is not displayed. Why would you want this? It's to allow text to avoid jumping around when adjust_spacing is True.

When the done tag is present, the line of dialogue is not added to the history buffer. If the nw tag is present, it should be before the done tag.:

g "Looks like they're{nw}{done} playing with their trebuchet again."
show trebuchet
g "Looks like they're{fast} playing with their trebuchet again."

clear

The clear text tag only makes sense on a line by itself in the NVL monologue mode. It does the same thing at the nvl clear statement without ending the block of text.

"""
Block 1 on page 1.

Block 2 on page 1.

{clear}

Block 1 on page 2.

etc.
"""

It's also possible to define custom text tags using Python.

Style Text Tags

Ren'Py supports text tags that access styles. These are text tags where the tag name is empty. In this case, the argument is taken to be the name of a style. For example, the {=mystyle} tag will access the mystyle style.

The text between the tag and the corresponding closing tag has the following properties set to those defined in the style:

antialias
font
size
bold
italic
underline
strikethrough
color
black_color
kerning

Non-English Languages

The default font for Ren'Py contains characters for English and many other languages. For size reasons, it doesn't contain the characters required to render other languages, including Chinese, Japanese, and Korean. In order to support these languages, a project must first change the fonts it uses.

Ren'Py should then support most world languages without further configuration. However, Korean can be written with or without spacing between words. Ren'Py has a special mode to support Korean with spaces, which can be enabled by setting:

define gui.language = "korean-with-spaces"

This can be changed from the default of "unicode" in gui.rpy.

Japanese has multiple rules for line breaking. We recommend starting with "japanese-normal", and moving to "japanese-loose" or "japanese-strict" for more or less break opportunities, respectively.

define gui.language = "japanese-loose"

Ideographic languages provide a large number of opportunities for line breaking. To enable a faster but less-accurate line-breaking algorithm, use:

define gui.language = "greedy"

The faster line-breaking algorithm is not be necessary unless the game is displaying huge amounts of text, such as in NVL-mode.

The line breaking algorithms can be further configured using the renpy.language_tailor() function.

Vertical Text

When the vertical style property is set, Ren'Py will produce vertically oriented text. The text is written top-to-bottom, right-to-left.

There are two text tags that interact with vertical text.

horiz: Includes horizontally-oriented text inside vertical text.

vert: Includes vertically-oriented text inside horizontal text. (This will not rotate the text to the vertical orientation.)

Note

If the font does not contain vertical text spacing information, Ren'Py will attempt to synthesize this information from horizontal text information. The spacing may not remain constant between Ren'Py releases.

Ruby Text

Ruby text (also known as furigana or interlinear annotations) is a way of placing small text above a character or word. There are several steps required for your game to support ruby text.

First, you must set up styles for the ruby text. The following style changes are required:

One of the line_leading or ruby_line_leading properties must be used to leave enough vertical space for the ruby text.
A new named style must be created. The properties of this style, such as size should be set in a fashion appropriate for ruby text.
The yoffset of the new style should be set, in order to move the ruby text above the baseline.
The ruby_style field of the text's style should be set to the newly-created style, for both dialogue and history window text.

For example:

style ruby_style is default:
    size 12
    yoffset -20
    color None # Use the same color as the parent text.

style say_dialogue:
    ruby_line_leading 12
    ruby_style style.ruby_style

style history_text:
    ruby_line_leading 12
    ruby_style style.ruby_style

(Use style.style_name to refer to a style for this purpose.)

Once Ren'Py has been configured, ruby text can be included in two way.

Lenticular brackets. Ruby text can be written by enclosing it full-width lenticular brackets (【】), with the full-width or half-width vertical line character (｜ or |) separating the bottom text from the top text. For example:

e "Ruby can be used for furigana (【東｜とう】 【京｜きょう】)."

e "It's also used for translations (【東京｜Tokyo】)."

Ruby text will only trigger if a vertical line is present. The left lenticular bracket can be quoted by doubling it. Lenticular ruby text may not contain other text tags.

The {rt} and {rb} text tags. The {rt} tag is used to mark one or more characters to be displayed as ruby text. If the ruby text is preceded by text enclosed in the {rb} tag, the ruby text is centered over that text. Otherwise, it is centered over the preceding character.

For example:

e "Ruby can be used for furigana (東{rt}とう{/rt} 京{rt}きょう{/rt})."

e "It's also used for translations ({rb}東京{/rb}{rt}Tokyo{/rt})."

It's the creator's responsibility to ensure that ruby text does not leave the boundaries of the text. It may be necessary to add leading or spaces to the left and right of the text to prevent these errors from occurring.

Ren'Py also supports alternate ruby text, which is a second kind of ruby top text. This is introduced with the {art} text tag (instead of {rt}), and the altruby_style property (instead of ruby_style).

Fonts

Ren'Py supports TrueType/OpenType fonts and collections, and Image-Based fonts.

A TrueType or OpenType font is specified by giving the name of the font file. The file must be present in the game directory or one of the archive files.

Ren'Py also supports TrueType/OpenType collections that define more than one font. When accessing a collection, use the 0-based font index, followed by an at-sign and the file name. For example, "0@font.ttc" is the first font in a collection, "1@font.ttc" the second, and so on.

When looking for a font files, if the file is not found, Ren'Py will search in the game/fonts directory. For example, when looking for "test.ttf", Ren'Py will first search for game/test.ttf, and then for game/fonts/test.ttf.

Font Replacement

The config.font_replacement_map variable is used to map fonts. The combination of font filename, boldness, and italics is mapped to a similar combination. This allows a font with proper italics to be used instead of the automatically-generated italics.

Once such mapping would be to replace the italic version of the Deja Vu Sans font with the official oblique version (You'll need to download the oblique font from the web):

init python:
    config.font_replacement_map["DejaVuSans.ttf", False, True] = ("DejaVuSans-Oblique.ttf", False, False)

This mapping can improve the look of italic text.

Font names and aliases

The config.font_name_map variable is used to give names or aliases to fonts. This allows two things : first, to make dialogues using the {font} tag shorter and easier to use, and second, to enable Font Groups to be used by these tags.

define config.font_name_map["jap"] = "electroharmonix.ttf"
define config.font_name_map["tjap"] = FontGroup().add("OrthodoxHerbertarian.ttf", "A", "Z").add("electroharmonix.ttf", None, None)

label yamato:
    e "Sorry, what does {font=jap}Black holes and revelations{/font} mean ?"
    y "You pronounce it {font=tjap}Black Holes And Revelations{/font}." # the capital letters appear in OrthodoxHerbertarian

Image-Based Fonts

Image based fonts can be registered by calling one of the following registration functions. Registering an image-based font requires the specification of a name, size, boldness, italicness, and underline. When all of these properties match the registered font, the registered font is used.

renpy.register_bmfont(name=None, size=None, bold=False, italics=False, underline=False, filename=None)

This registers a BMFont with the given details. Please note that size, bold, italic, and underline are all advisory (used for matching), and do not change the appearance of the font.

Please see the BMFont home page for the tool that creates BMFonts. Ren'Py expects that the filename parameter will be to a file in the BMFont text format, that describes a 32-bit font. The Alpha channel should contain the font information, while the Red, Green, and Blue channels should be set to one. The image files, kerning, and other control information is read out of the BMFont file.

We recommend including Latin and General Punctuation as part of your BMFont, to ensure all of the Ren'Py interface can render.

name: The name of the font being registered, a string.
size: The size of the font being registered, an integer.
bold: The boldness of the font being registered, a boolean.
italics: The italicness of the font being registered, a boolean.
underline: An ignored parameter.
filename: The file containing BMFont control information.

renpy.register_mudgefont(name=None, size=None, bold=False, italics=False, underline=False, filename=None, xml=None, spacewidth=10, default_kern=0, kerns={})

This registers a MudgeFont with the given details. Please note that size, bold, italic, and underline are all advisory (used for matching), and do not change the appearance of the font.

Please see the MudgeFont home page for the tool that creates MudgeFonts. Ren'Py assumes that character codes found in the MudgeFont xml file are unicode character numbers, and ignores negative character codes.

name: The name of the font being registered, a string.
size: The size of the font being registered, an integer.
bold: The boldness of the font being registered, a boolean.
italics: The italicness of the font being registered, a boolean.
underline: An ignored parameter.
filename: The file containing the MudgeFont image, a string. The image is usually a TGA file, but could be a PNG or other format that Ren'Py supports.
xml: The xml file containing information generated by the MudgeFont tool.
spacewidth: The width of a space character, an integer in pixels.
default_kern: The default kern spacing between characters, in pixels.
kerns: A map from two-character strings to the kern that should be used between those characters.

renpy.register_sfont(name=None, size=None, bold=False, italics=False, underline=False, filename=None, spacewidth=10, baseline=None, default_kern=0, kerns={}, charset='!"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~')

This registers an SFont with the given details. Please note that size, bold, italic, and underline are all advisory (used for matching), and do not change the appearance of the font.

More information about SFont.

name: The name of the font being registered, a string.
size: The size of the font being registered, an integer.
bold: The boldness of the font being registered, a boolean.
italics: The italicness of the font being registered, a boolean.
underline: An ignored parameter.
filename: The file containing the sfont image, a string.
spacewidth: The width of a space character, an integer in pixels.
baseline: The distance from the top of the font to the baseline (the invisible line letters sit on), an integer in pixels. If this font is mixed with other fonts, their baselines will be aligned. Negative values indicate distance from the bottom of the font instead, and None means the baseline equals the height (i.e., is at the very bottom of the font).
default_kern: The default kern spacing between characters, in pixels.
kerns: A map from two-character strings to the kern that should be used between those characters.
charset: The character set of the font. A string containing characters in the order in which they are found in the image. The default character set for a SFont is

! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ?
@ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _
` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~

As BMFont is the most complete of the three image font formats Ren'Py supports, it's the one recommended for new projects. An example of BMFont use is:

init python:
    renpy.register_bmfont("bmfont", 22, filename="bmfont.fnt")

define ebf = Character('Eileen', what_font="bmfont", what_size=22)

label demo_bmfont:

    ebf "Finally, Ren'Py supports BMFonts."

Font Groups

When creating a multilingual game, it may not be possible to find a single font that covers every writing system the game use while projecting the the mood the creator intends. To support this, Ren'Py supports font groups that can take characters from two or more fonts and combine them into a single font.

To create a font group, create a FontGroup object and call the .add method on it once or more. A FontGroup can be used wherever a font name can be used. The add method takes the start and end of a range of Unicode character points, and the first range to cover a point is used.

For example:

style default:
    font FontGroup().add("english.ttf", 0x0020, 0x007f).add("japanese.ttf", 0x0000, 0xffff)

class FontGroup

A group of fonts that can be used as a single font.

add(font, start, end, target=None, target_increment=False)

Associates a range of characters with a font.

start: The start of the range. This may be a single-character string, or an integer giving a unicode code point. If start is None, then the font is used as the default.
end: The end of the range. This may be a single-character string, or an integer giving a unicode code point. This is ignored if start is None.
target: If given, associates the given range of characters with specific characters from the given font, depending on target_increment. This may be a single-character string, or an integer giving a unicode code point. This is ignored if the character had already been added.
target_increment: If True, the [start, end] range is mapped to the [target, target+end-start] range. If False, every character from the range is associated with the target character.

When multiple .add() calls include the same character, the first call takes precedence.

This returns the FontGroup, so that multiple calls to .add() can be chained together.

remap(cha, target)

Remaps one or a set of characters to a single target character.

cha: The character or characters to remap. This may be a single-character string, or an integer giving a unicode code point, or an iterable of either.
target: The character to remap to. This may be a single-character string, or an integer giving a unicode code point.

Any given character having already been remapped (either with add or with remap) will be ignored. However, if the FontGroup has no default font, any given character must have been previously added.

This method also returns the FontGroup, for the same reasons.

Note that while FontGroups can be given a name using config.font_name_map, a FontGroup only takes filepaths as fonts, and does not recognize names or aliases defined using that variable.

Text Displayables

Text can also be used as a displayable, which allows you to apply transforms to text, displaying it as if it was an image and moving it around the screen.

class ParameterizedText(style='default', **properties)

This is a displayable that can be shown with an additional string parameter, which then shows that string as if it was an image. This is usually used as part of the pre-defined text image.

For example, one can do:

show text "Hello, World" at truecenter
with dissolve
pause 1
hide text
with dissolve

You can use ParameterizedText directly to define similar images with different style properties. For example, one can write:

image top_text = ParameterizedText(xalign=0.5, yalign=0.0)

label start:
    show top_text "This text is shown at the center-top of the screen"

class Text(text, slow=None, scope=None, substitute=None, slow_done=None, *, tokenized=False, **properties)

A displayable that displays text on the screen.

text: The text to display on the screen. This may be a string, or a list of strings and displayables.
slow: Determines if the text is displayed slowly, being typed out one character at the time. If None, slow text mode is determined by the slow_cps style property. Otherwise, the truth value of this parameter determines if slow text mode is used.
scope: If not None, this should be a dictionary that provides an additional scope for text interpolation to occur in.
substitute: If true, text interpolation occurs. If false, it will not occur. If None, they are controlled by config.new_substitutions.
slow_done: If not None, and if slow text mode is enabled (see the slow parameter), this is a function or callable which is called with no arguments when the text finishes displaying.
tokenized: If true, text is expected to be a list of tokens, rather than a string. The tokens are introduced in the Custom Text Tags page.
**properties: Like other Displayables, Text takes style properties, including (among many others) the mipmap property.

Text Utility Functions

renpy.filter_text_tags(s, allow=None, deny=None)

Returns a copy of s with the text tags filtered. Exactly one of the allow and deny keyword arguments must be given.

allow: A set of tags that are allowed. If a tag is not in this list, it is removed.
deny: A set of tags that are denied. If a tag is not in this list, it is kept in the string.

renpy.substitute(s, scope=None, translate=True)

Applies translation and new-style formatting to the string s.

scope: If not None, a scope which is used in formatting, in addition to the default store.
translate: Determines if translation occurs.

Returns the translated and formatted string.

renpy.transform_text(s, function)

Transforms s, while leaving text tags and interpolation the same.

function: A function that is called with strings corresponding to runs of text, and should return a second string that replaces that run of text.

init python:
    def upper(s):
        return s.upper()

$ upper_string = renpy.transform_text("{b}Not Upper{/b}", upper)

Slow Text Concerns

Ren'Py allows the creator or user to indicate that text should be displayed slowly. In this case, Ren'Py will render the text to a texture, and then draw rectangles from the texture to the screen.

Unfortunately, this means that it's possible to get rendering artifacts when characters overlap. To minimize these rendering artifacts, ensure that the line_leading and line_spacing properties are large enough that lines do not overlap. If the bottoms of characters on the first line are clipped, especially if line_spacing is negative, consider increasing line_overlap_split.

Horizontal artifacts are also possible when characters are kerned together, but these artifacts are less severe, as they exist for only a single frame.

Artifacts aren't a problem for static text, like the text in menus and other parts of the user interface.

Text Overflow Logging

Ren'Py can log cases where text expands outside of the area allocated for it. To enable text overflow logging, the following steps are necessary.

Set the config.debug_text_overflow variable to true.
Set the xmaximum and ymaximum style properties on either the Text displayable, or a container enclosing it.
Run the game.

Whenever text is displayed that overflows the available area, Ren'Py will log an error to the text_overflow.txt file.

Variable Fonts

Ren'Py has support for OpenType variable fonts. These fonts may support multiple axes, such as weight and width, and based on the value of these axes, the display of the font can change. Variable fonts may also have named instances, which provide different values for the axes. For example, the "bold" named instance will likely provide the bold version of a font, while "regular" will be a more normal version.

Variable fonts require the use of the harfbuzz text shaper, controlled by the shaper style property. Once this is set, the instance property selects a named instance, while the axis property sets the value of one or more axes.

These can also be set in the gui system. For example one can use:

define gui.text_font = "nunito.ttf"
define gui.text_instance = "light"
define gui.text_axis = {"width" : 125}

Will start with the light instance, and make it wide.

If no instance is given, Ren'Py will default to "regular" for non-bold text, and "bold" for bold text.

There are two text tags that support the use of variable fonts.

instance

The instance tag changes the instance that is being used. For example:

"This is {instance=heavy}heavy{/instance} text."

When the instance tag is used, the axis properties are overridden.

axis

The axis tag changes the value of one or more axes. For example:

"This is {axis:width=125}wide{/axis} text."

The axis tag can be used multiple times, and the values will be combined.

"This is {axis:width=125}{axis:weight=200}wide and bold{/axis}{/axis} text."

The value of on the right side of the equals is expected to be a floating point number.

To get the available instances and axes of a font, use the renpy.variable_font_info() function. This function is intended to be called from the console. To access it, hit shift+O to open the console, and then type:

renpy.variable_font_info("nunito.ttf")

This will display information for the nunito.ttf font.

renpy.variable_font_info(font)

Returns information about a variable font, or None if the font is not variable.

font: The filename containing the font.

The return value is an object with the following fields:

instance

A dictionary where the keys are the names of the named instances of the font. (For example, 'light', 'regular', 'bold', and 'heavy'.) The values can be ignored.

axis

A dictionary that maps the names of the axes of the font to objects with the following fields:

minimum: The minimum value of the axis.
default: The default value of the axis.
maximum: The maximum value of the axis.

The object returned by this function and the data inside it should not be changed.

This function may only be called after the Ren'Py display has initialized, and is intended to be called from the console, where it will print in a human-readable form.