diff --git a/CMakeLists.txt b/CMakeLists.txt
index 3fc6567..c4d6277 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -14,6 +14,7 @@ set(CMAKE_LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/lib)
 add_executable(cpp-workspace-app
     src/main.cpp
     src/ollama_client.cpp
+    src/progress_spinner.cpp
 )
 
 find_package(fmt REQUIRED)
diff --git a/src/main.cpp b/src/main.cpp
index d8de3f1..fb7a29c 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -2,38 +2,38 @@
 #include <iostream>
 
 int main(int argc, const char* argv[]) {
-    std::string model_arg = "gemma3n:latest";
-    std::string prompt_arg = "Hey! What's up!";
-    bool stream_arg = false;
+    std::string model = "gemma3n:latest";
+    std::string prompt = "Hey! What's up!";
+    bool stream = false;
 
     for (int i = 1; i < argc; i += 1) {
         std::string arg = argv[i];
         if (arg.starts_with("--model=") == 0) {
-            model_arg = arg.substr(8);
+            model = arg.substr(8);
         } else if (arg.starts_with("--prompt=") == 0) {
-            prompt_arg = arg.substr(9);
+            prompt = arg.substr(9);
         } else if (arg.starts_with("--stream=") == 0) {
-            stream_arg = arg.substr(9) == "true" ? true : false;
+            stream = arg.substr(9) == "true" ? true : false;
         }
     }
 
     printf("\n");
-    printf("model_arg = %s\n", model_arg.c_str());
-    printf("prompt_arg = %s\n", prompt_arg.c_str());
-    printf("stream_arg = %s\n", stream_arg ? "true" : "false");
+    printf("🛈 model = %s\n", model.c_str());
+    printf("🛈 prompt = %s\n", prompt.c_str());
+    printf("🛈 stream = %s\n", stream ? "true" : "false");
     printf("\n");
 
-    if (prompt_arg.length() == 0) {
+    if (prompt.length() == 0) {
         printf("Enter your prompt: ");
-        getline(std::cin, prompt_arg);
+        getline(std::cin, prompt);
     }
 
     OllamaClient client = OllamaClient();
 
     struct OllamaPromptConfig config;
-    config.model = model_arg;
-    config.prompt = prompt_arg;
-    config.stream = stream_arg;
+    config.model = model;
+    config.prompt = prompt;
+    config.stream = stream;
 
     client.curl_request(config);
 
diff --git a/src/ollama_client.cpp b/src/ollama_client.cpp
index d140391..78c1bce 100644
--- a/src/ollama_client.cpp
+++ b/src/ollama_client.cpp
@@ -1,8 +1,10 @@
 #include "ollama_client.hpp"
+#include <algorithm>
 #include <curl/curl.h>
 #include <iostream>
 #include <nlohmann/json.hpp>
 #include <sstream>
+#include <string>
 
 using nlohmann::json_abi_v3_11_2::json;
 
@@ -30,7 +32,7 @@ void OllamaClient::check_ollama() {
                         "https://docs.ollama.com/linux\n");
         exit(0);
     } else {
-        printf("[ok] Ollama binary exists.\n");
+        printf("✅ Ollama binary exists.\n");
     }
 };
 
@@ -41,15 +43,31 @@ void OllamaClient::init_curl() {
         fprintf(stderr, "Failed to initialize CURL.\n");
         exit(1);
     } else {
-        printf("[ok] CURL initialized.\n");
+        printf("✅ CURL initialized.\n");
     }
 }
 
+// Trims leading whitespace and trailing whitespace from a json response.
+std::string OllamaClient::trim_response(std::string response) {
+    response.erase(response.begin(),
+                   std::find_if(response.begin(), response.end(),
+                                [](unsigned char ch) { return !std::isspace(ch); }));
+
+    response.erase(std::find_if(response.rbegin(), response.rend(),
+                                [](unsigned char ch) { return !std::isspace(ch); })
+                       .base(),
+                   response.end());
+    return response;
+}
+
 // Process a response.
 void OllamaClient::process_response(std::string response) {
     try {
         json response_json = json::parse(response);
-        std::cout << "Response: " << response_json["response"].get<std::string>() << std::endl;
+
+        std::cout << "🦙 llama:\n"
+                  << this->trim_response(response_json["response"].get<std::string>()) << "\n---\n"
+                  << std::endl;
     } catch (const json::exception& err) {
         fprintf(stderr, "JSON parse error: %s\n", err.what());
     }
@@ -97,6 +115,9 @@ void OllamaClient::curl_request(OllamaPromptConfig config) {
     curl_easy_setopt(this->curl, CURLOPT_WRITEFUNCTION, WriteCallback);
     curl_easy_setopt(this->curl, CURLOPT_WRITEDATA, &response);
 
+    curl_easy_setopt(this->curl, CURLOPT_XFERINFOFUNCTION, ProgressSpinner::progress_callback);
+    curl_easy_setopt(this->curl, CURLOPT_NOPROGRESS, 0L);
+
     this->headers = curl_slist_append(this->headers, "Content-Type: application/json");
     curl_easy_setopt(this->curl, CURLOPT_HTTPHEADER, this->headers);
 
diff --git a/src/ollama_client.hpp b/src/ollama_client.hpp
index 04f84f6..17276c5 100644
--- a/src/ollama_client.hpp
+++ b/src/ollama_client.hpp
@@ -1,3 +1,4 @@
+#include "progress_spinner.hpp"
 #include <curl/curl.h>
 #include <iostream>
 
@@ -12,9 +13,13 @@ class OllamaClient {
     CURL* curl;
     std::string url;
     struct curl_slist* headers;
+    ProgressSpinner spinner{ProgressSpinner()};
 
     void check_ollama();
     void init_curl();
+
+    std::string trim_response(std::string response);
+
     void process_response(std::string response);
     void process_response_stream(std::string response);
 
diff --git a/src/progress_spinner.cpp b/src/progress_spinner.cpp
new file mode 100644
index 0000000..6f216f4
--- /dev/null
+++ b/src/progress_spinner.cpp
@@ -0,0 +1,73 @@
+#include "progress_spinner.hpp"
+#include <atomic>
+#include <chrono>
+#include <curl/curl.h>
+#include <iostream>
+#include <thread>
+
+ProgressSpinner::ProgressSpinner() {
+    this->active = false;
+}
+
+ProgressSpinner::ProgressSpinner(bool activate) {
+    if (activate == true) {
+        this->active = activate;
+        start();
+    }
+}
+
+ProgressSpinner::~ProgressSpinner() {
+    stop();
+}
+
+// Starts the progress spinner.
+void ProgressSpinner::start() {
+    active = true;
+    spinner_thread = std::thread([this]() {
+        const char spinner_chars[] = {'|', '/', '-', '\\'};
+        int i = 0;
+        while (active) {
+            std::cout << "\r" << spinner_chars[i++ % 4] << " Loading..." << std::flush;
+            std::this_thread::sleep_for(std::chrono::milliseconds(100));
+        }
+        std::cout << "\r✓ Complete\n" << std::endl;
+    });
+}
+
+// Stops the progress spinner.
+void ProgressSpinner::stop() {
+    active = false;
+    if (spinner_thread.joinable()) {
+        spinner_thread.join();
+    }
+}
+
+/**
+ * CURL progress callback.
+ *
+ * @example
+ * ```cpp
+ * CURL* curl = curl_easy_init();
+ * curl_easy_setopt(this->curl, CURLOPT_XFERINFOFUNCTION, ProgressSpinner::progress_callback);
+ * curl_easy_setopt(this->curl, CURLOPT_NOPROGRESS, 0L);
+ * curl_easy_perform(curl);
+ * ```
+ */
+int ProgressSpinner::progress_callback(void* clientp, curl_off_t dltotal, curl_off_t dlnow,
+                                       curl_off_t ultotal, curl_off_t ulnow) {
+    if (dltotal > 0) {
+        // If size is known, show progress bar
+        int progress = (dlnow * 100) / dltotal;
+        if (progress == 100) {
+            std::cout << "\r✓ Complete  " << std::flush;
+        } else {
+            std::cout << "\rProgress: " << progress << "% " << std::flush;
+        }
+    } else {
+        // Indeterminate - show spinner
+        const char spinner_chars[] = {'|', '/', '-', '\\'};
+        static int i = 0;
+        std::cout << "\r" << spinner_chars[i++ % 4] << " Loading..." << std::flush;
+    }
+    return 0;
+}
diff --git a/src/progress_spinner.hpp b/src/progress_spinner.hpp
new file mode 100644
index 0000000..5857b8d
--- /dev/null
+++ b/src/progress_spinner.hpp
@@ -0,0 +1,39 @@
+#pragma once
+
+#ifndef PROGRESS_SPINNER_HPP
+#define PROGRESS_SPINNER_HPP
+
+#include <atomic>
+#include <curl/curl.h>
+#include <iostream>
+#include <thread>
+
+/**
+ * Progress spinner for asynchronous operations.
+ *
+ * @example manual mode
+ * ```cpp
+ * ProgressSpinner spinner = ProgressSpinner();
+ * spinner.start();
+ * curl_easy_perform(curl_handle);
+ * spinner.stop();
+ * ```
+ */
+class ProgressSpinner {
+  private:
+    std::atomic<bool> active{false};
+    std::thread spinner_thread;
+
+  public:
+    ProgressSpinner();
+    explicit ProgressSpinner(bool activate);
+    ~ProgressSpinner();
+
+    void start();
+    void stop();
+
+    static int progress_callback(void* clientp, curl_off_t dltotal, curl_off_t dlnow,
+                                 curl_off_t ultotal, curl_off_t ulnow);
+};
+
+#endif