diff --git a/src/modules/ttsSettings.ts b/src/modules/ttsSettings.ts
index ad9be38..2dd6991 100644
--- a/src/modules/ttsSettings.ts
+++ b/src/modules/ttsSettings.ts
@@ -1,4 +1,5 @@
 import { join } from "node:path";
+import { AttachmentBuilder } from "discord.js";
 import { respond } from "../audio/AudioService.js";
 import type { Module } from "./types.js";
 
@@ -36,5 +37,32 @@ export const ttsSettings: Module = ({ audio, commands, tts, t, rootDir }) => {
     }
   });
 
+  commands.register("lsvoices", (args, message) => {
+    if (args.length > 1) {
+      respond(audio, sysmsg, message, t("TOO_MANY_ARGUMENTS"));
+      return;
+    }
+    const sections: string[] = [];
+    for (const name of tts.list()) {
+      const engine = tts.get(name);
+      if (!engine) continue;
+      const lines = [`=== ${name} — ${engine.longName} ===`, `default: ${engine.getDefaultVoice()}`];
+      if (engine.isFreeformVoice()) {
+        lines.push(`(accepts any voice string — see ${engine.longName} documentation)`);
+      } else {
+        const voices = engine.listVoices();
+        lines.push(`voices (${voices.length}):`);
+        for (const v of voices) lines.push(`  ${v}`);
+      }
+      sections.push(lines.join("\n"));
+    }
+    const body = sections.join("\n\n") + "\n";
+    const file = new AttachmentBuilder(Buffer.from(body, "utf8"), { name: "voices.txt" });
+    void message.reply({
+      content: `Available TTS engines and voices (${tts.list().length} engines):`,
+      files: [file],
+    });
+  });
+
   commands.register("flush", () => audio.queue.flush());
 };
diff --git a/src/tts/BaseEngine.ts b/src/tts/BaseEngine.ts
index 58dd506..f435289 100644
--- a/src/tts/BaseEngine.ts
+++ b/src/tts/BaseEngine.ts
@@ -37,6 +37,16 @@ export abstract class BaseEngine {
     return this.voices[voice] != null;
   }
 
+  /** Returns sorted lowercase voice names known to this engine. Empty if the engine accepts any voice. */
+  listVoices(): string[] {
+    return Object.keys(this.voices).sort();
+  }
+
+  /** True when this engine has no static voice table and accepts any voice string. */
+  isFreeformVoice(): boolean {
+    return Object.keys(this.voices).length === 0;
+  }
+
   /** Default implementation: subclass should override either this or getSpeechFile. */
   async getSpeech(_text: string, _voice?: string, _params?: VoiceParams): Promise<Response> {
     throw new Error(`${this.shortName}: getSpeech not implemented`);