diff --git a/NAMESPACE b/NAMESPACE
index 1db4b1b..ab4ef2e 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -14,6 +14,7 @@ export(generate)
 export(image_encode_base64)
 export(insert_message)
 export(list_models)
+export(ohelp)
 export(prepend_message)
 export(pull)
 export(resp_process)
diff --git a/R/ollama.R b/R/ollama.R
index f8f0d81..9198e2d 100644
--- a/R/ollama.R
+++ b/R/ollama.R
@@ -568,3 +568,44 @@ embeddings <- function(model, prompt, normalize = TRUE, keep_alive = "5m", endpo
 
 
 
+#' Chat with a model in real-time in R console.
+#'
+#' @param model A character string of the model name such as "llama3". Defaults to "codegemma:7b" which is a decent coding model as of 2024-07-27.
+#' @param ... Additional options. No options are currently available at this time.
+#'
+#' @return Does not return anything. It prints the conversation in the console.
+#' @export
+#'
+#' @examplesIf test_connection()$status_code == 200
+#' ohelp(first_prompt = "quit")
+#' # regular usage: ohelp()
+ohelp <- function(model = "codegemma:7b", ...) {
+
+    cat("Say something or type /q to quit or end the conversation.\n\n")
+
+    n_messages <- 0
+    opts <- list(...)
+    if (length(opts) > 0) {
+        if (opts$first_prompt == "quit") {
+            prompt <- "/q"
+        }
+    } else {
+        prompt <- readline()
+    }
+
+    while (prompt != "/q") {
+        if (n_messages == 0) {
+            messages <- create_message(prompt, role = 'user')
+        } else {
+            messages <- append(messages, create_message(prompt, role = 'user'))
+        }
+        n_messages <- n_messages + 1
+        response <- chat(model, messages = messages, output = 'text', stream = TRUE)
+        messages <- append_message(response, "assistant", messages)
+        n_messages <- n_messages + 1
+        prompt <- readline()
+    }
+
+    cat("Goodbye!\n")
+
+}
diff --git a/_pkgdown.yml b/_pkgdown.yml
index ff2af29..099d494 100644
--- a/_pkgdown.yml
+++ b/_pkgdown.yml
@@ -14,6 +14,7 @@ reference:
       - pull
       - embed
       - embeddings
+      - ohelp
       - test_connection
       - create_request
 
diff --git a/man/image_encode_base64.Rd b/man/image_encode_base64.Rd
index 2415b0d..5e036e3 100644
--- a/man/image_encode_base64.Rd
+++ b/man/image_encode_base64.Rd
@@ -16,6 +16,6 @@ A base64 encoded string.
 Read image file and encode it to base64.
 }
 \examples{
-image_path <- file.path(system.file('extdata', package = "ollamar"), "image1.png")
+image_path <- file.path(system.file("extdata", package = "ollamar"), "image1.png")
 image_encode_base64(image_path)
 }
diff --git a/man/ohelp.Rd b/man/ohelp.Rd
new file mode 100644
index 0000000..ab83b40
--- /dev/null
+++ b/man/ohelp.Rd
@@ -0,0 +1,25 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/ollama.R
+\name{ohelp}
+\alias{ohelp}
+\title{Chat with a model in real-time in R console.}
+\usage{
+ohelp(model = "codegemma:7b", ...)
+}
+\arguments{
+\item{model}{A character string of the model name such as "llama3". Defaults to "codegemma:7b" which is a decent coding model as of 2024-07-27.}
+
+\item{...}{Additional options. No options are currently available at this time.}
+}
+\value{
+Does not return anything. It prints the conversation in the console.
+}
+\description{
+Chat with a model in real-time in R console.
+}
+\examples{
+\dontshow{if (test_connection()$status_code == 200) (if (getRversion() >= "3.4") withAutoprint else force)(\{ # examplesIf}
+ohelp(first_prompt = "quit")
+# regular usage: ohelp()
+\dontshow{\}) # examplesIf}
+}