Skip to content

DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.

fix(voice): hide unsupported DashScope 8k STT option

Bingxi Zhao (Frank)7adef0b607ae
diff --git a/deeptutor/services/voice/adapters/dashscope.py b/deeptutor/services/voice/adapters/dashscope.py
index 3f7b0fb58..b5b034eb6 100644
--- a/deeptutor/services/voice/adapters/dashscope.py
+++ b/deeptutor/services/voice/adapters/dashscope.py
@@ -135,6 +135,11 @@ class DashScopeSTTAdapter(BaseSTTAdapter):
) -> str:
if not audio:
raise VoiceProviderError("No audio data to transcribe.")
+ if config.model in {"paraformer-realtime-8k-v1", "paraformer-realtime-8k-v2"}:
+ raise VoiceProviderError(
+ "DashScope 8k realtime models require 8000 Hz audio; "
+ "select paraformer-realtime-v2 for the 16000 Hz voice adapter."
+ )
wav_audio = await self._prepare_wav(audio, filename, content_type)
if not wav_audio:
raise VoiceProviderError("Audio conversion returned an empty file.")
diff --git a/deeptutor/services/voice/options.py b/deeptutor/services/voice/options.py
index 29ddf29fd..84ab11779 100644
--- a/deeptutor/services/voice/options.py
+++ b/deeptutor/services/voice/options.py
@@ -201,7 +201,7 @@ def voice_options(provider: str, service: str) -> dict:
"openrouter": ["openai/whisper-large-v3"],
"groq": ["whisper-large-v3-turbo", "whisper-large-v3"],
"siliconflow": ["FunAudioLLM/SenseVoiceSmall"],
- "dashscope": ["paraformer-realtime-v2", "paraformer-realtime-8k-v2"],
+ "dashscope": ["paraformer-realtime-v2"],
"volcengine_speech": ["bigmodel"],
}.get(provider, [])
languages = VOLC_LANGUAGES if provider == "volcengine_speech" else ISO_LANGUAGES
diff --git a/tests/services/test_voice.py b/tests/services/test_voice.py
index 115624c2c..77e40d306 100644
--- a/tests/services/test_voice.py
+++ b/tests/services/test_voice.py
@@ -31,12 +31,14 @@ from deeptutor.services.voice.adapters.openai_compat import (
OpenRouterTTSAdapter,
)
from deeptutor.services.voice.base import (
+ VoiceProviderError,
build_auth_headers,
join_audio_path,
normalize_stt_content_type,
strip_markdown_for_speech,
)
from deeptutor.services.voice.config import STTConfig, TTSConfig
+from deeptutor.services.voice.options import voice_options
def _capture_post(monkeypatch: pytest.MonkeyPatch, response: httpx.Response) -> dict[str, Any]:
@@ -488,6 +490,18 @@ def test_dashscope_stt_url_and_errors() -> None:
assert adapter._sentence_texts({"sentence": {"text": "single"}}) == ["single"]
+def test_dashscope_stt_options_only_offer_supported_sample_rate() -> None:
+ models = voice_options("dashscope", "stt")["models"]
+ assert [model["id"] for model in models] == ["paraformer-realtime-v2"]
+
+
+async def test_dashscope_stt_rejects_8k_model_before_audio_conversion() -> None:
+ config = STTConfig(model="paraformer-realtime-8k-v2", api_key="dash-key")
+ with pytest.raises(VoiceProviderError, match="require 8000 Hz audio"):
+ await DashScopeSTTAdapter().transcribe(b"audio", config)
+
+
@pytest.mark.asyncio
async def test_stt_adapter_openrouter_base64(monkeypatch: pytest.MonkeyPatch) -> None:
resp = httpx.Response(200, json={"text": "from base64"})