Skip to content

These functions are used to prepare image URLs and files for input to the chatbot. The content_image_url() function is used to provide a URL to an image, while content_image_file() is used to provide the image data itself.

Usage

content_image_url(url, detail = c("auto", "low", "high"))

content_image_file(path, content_type = "auto", resize = "low")

content_image_plot(width = 768, height = 768)

Arguments

url

The URL of the image to include in the chat input. Can be a data: URL or a regular URL. Valid image types are PNG, JPEG, WebP, and non-animated GIF.

detail

The detail setting for this image. Can be "auto", "low", or "high".

path

The path to the image file to include in the chat input. Valid file extensions are .png, .jpeg, .jpg, .webp, and (non-animated) .gif.

content_type

The content type of the image (e.g. image/png). If "auto", the content type is inferred from the file extension.

resize

If "low", resize images to fit within 512x512. If "high", resize to fit within 2000x768 or 768x2000. (See the OpenAI docs for more on why these specific sizes are used.) If "none", do not resize.

You can also pass a custom string to resize the image to a specific size, e.g. "200x200" to resize to 200x200 pixels while preserving aspect ratio. Append > to resize only if the image is larger than the specified size, and ! to ignore aspect ratio (e.g. "300x200>!").

All values other than none require the magick package.

width, height

Width and height in pixels.

Value

An input object suitable for including in the ... parameter of the chat(), stream(), chat_async(), or stream_async() methods.

Examples

chat <- chat_openai(echo = TRUE)
#> Using model = "gpt-4o".
chat$chat(
  "What do you see in these images?",
  content_image_url("https://www.r-project.org/Rlogo.png"),
  content_image_file(system.file("httr2.png", package = "ellmer"))
)
#> The first image is the logo of the R programming language, which 
#> consists of a gray letter "R" inside a gray oval.
#> 
#> The second image is a hexagonal sticker for the "httr2" R package. It 
#> features a stylized illustration of a baseball player swinging a bat 
#> and the text "httr2" in a script font.

plot(waiting ~ eruptions, data = faithful)

chat <- chat_openai(echo = TRUE)
#> Using model = "gpt-4o".
chat$chat(
  "Describe this plot in one paragraph, as suitable for inclusion in
   alt-text. You should briefly describe the plot type, the axes, and
   2-5 major visual patterns.",
   content_image_plot()
)
#> This plot is a line graph depicting the sinusoidal functions sine 
#> (sin) and cosine (cos) over the interval from 0 to 2π on the x-axis, 
#> with values ranging from -1 to 1 on the y-axis. The sine curve, shown 
#> in blue, begins at 0, peaks at 1 at π/2, crosses zero at π, reaches -1
#> at 3π/2, and returns to 0 at 2π. The cosine curve, in orange, starts 
#> at 1, crosses zero at π/2, reaches -1 at π, returns to zero at 3π/2, 
#> and finally returns to 1 at 2π. The two curves are out of phase by 
#> π/2, showing complementary peaks and troughs.