diff --git a/apps/cli/src/git.zig b/apps/cli/src/git.zig
index 5999590..eafea59 100644
--- a/apps/cli/src/git.zig
+++ b/apps/cli/src/git.zig
@@ -7,6 +7,11 @@ pub fn getStagedChanges(allocator: std.mem.Allocator) ![]u8 {
         "--cached",
     } });
 
+    if (result.stdout.len == 0) {
+        try std.io.getStdOut().writer().print("No changes to commit\n", .{});
+        std.process.exit(0);
+    }
+
     return result.stdout;
 }
 
diff --git a/apps/cli/src/main.zig b/apps/cli/src/main.zig
index 4ecc3a6..287fdd0 100644
--- a/apps/cli/src/main.zig
+++ b/apps/cli/src/main.zig
@@ -8,7 +8,7 @@ var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
 const allocator = arena.allocator();
 
 var config = struct {
-    model: []const u8 = "gpt-4-turbo-preview",
+    model: []const u8 = openai.Model.Default.toString(),
 }{};
 
 var model = cli.Option{
@@ -19,11 +19,6 @@ var model = cli.Option{
 
 fn generate_handler() !void {
     const diff = try git.getStagedChanges(allocator);
-    if (diff.len == 0) {
-        try std.io.getStdOut().writer().print("No changes to commit\n", .{});
-        return;
-    }
-
     const response = try openai.getCompletion(allocator, diff, config.model);
 
     try std.io.getStdOut().writer().print("{s}\n", .{response.choices[0].message.content});
@@ -31,11 +26,6 @@ fn generate_handler() !void {
 
 fn commit_handler() !void {
     const diff = try git.getStagedChanges(allocator);
-    if (diff.len == 0) {
-        try std.io.getStdOut().writer().print("No changes to commit\n", .{});
-        return;
-    }
-
     const response = try openai.getCompletion(allocator, diff, config.model);
     const message = response.choices[0].message.content;
 
diff --git a/apps/cli/src/openai.zig b/apps/cli/src/openai.zig
index b2371d8..d56d8c3 100644
--- a/apps/cli/src/openai.zig
+++ b/apps/cli/src/openai.zig
@@ -20,6 +20,24 @@ const CompletionResponse = struct {
     choices: []Choice,
 };
 
+pub const Model = enum(u2) {
+    Fast,
+    Default,
+
+    /// Returns the string representation of the enum value
+    pub fn toString(self: Model) []const u8 {
+        return switch (self) {
+            .Fast => "gpt-3.5-turbo",
+            .Default => "gpt-4-turbo-preview",
+        };
+    }
+
+    /// Checks whether the provided model is supported
+    pub fn isSupported(model: []const u8) bool {
+        return std.mem.eql(u8, model, Model.Fast.toString()) or std.mem.eql(u8, model, Model.Default.toString());
+    }
+};
+
 const SYSTEM_MESSAGE =
     \\ You are a helpful coding assistant responsible for generating fitting commit messages.
     \\ You will be provided a git diff or code snippet and you are expected to provide a suitable commit message.
@@ -48,6 +66,11 @@ const SYSTEM_MESSAGE =
 ;
 
 pub fn getCompletion(allocator: std.mem.Allocator, prompt: []const u8, model: []const u8) !CompletionResponse {
+    if (!Model.isSupported(model)) {
+        std.debug.print("Unsupported model provided. If you believe the model should be supported, report an issue at https://github.com/segersniels/genmoji/issues.\n", .{});
+        std.process.exit(1);
+    }
+
     var env = try std.process.getEnvMap(allocator);
     defer env.deinit();