diff --git a/src/models/higgs_audio_tts/generator.cpp b/src/models/higgs_audio_tts/generator.cpp
index 2820067e..c97b34ed 100644
--- a/src/models/higgs_audio_tts/generator.cpp
+++ b/src/models/higgs_audio_tts/generator.cpp
@@ -508,7 +508,11 @@ HiggsGenerationResult HiggsGenerator::generate(const HiggsGenerationRequest & re
     engine::debug::timing_log_scalar("higgs_audio_tts.generator.decode_ms",
                                      engine::debug::elapsed_ms(decode_start, Clock::now()));
     if (!state.generation_done) {
-        throw std::runtime_error("Higgs TTS generation reached max_tokens before EOC");
+        throw std::runtime_error(
+            "Higgs TTS generation reached max_tokens (" + std::to_string(request.options.max_tokens) +
+            ") before EOC for this text chunk; raise it with --max-tokens on the CLI or "
+            "the \"max_tokens\" request option on the server, or lower --text-chunk-size / "
+            "\"text_chunk_size\" so each chunk needs fewer generated frames");
     }
 
     result.raw_codes = reverse_higgs_delay_pattern(