DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.
Python ai-agentsai-tutorclawbotcli-tooldeepresearchinteractive-learninglearningmulti-agent-systemsrag
fix(voice): use realtime DashScope STT model ids
wumajiehechuan-labcdc36c31d355
diff --git a/deeptutor/services/config/provider_runtime.py b/deeptutor/services/config/provider_runtime.pyindex 288f6876a..4b7778b57 100644--- a/deeptutor/services/config/provider_runtime.py+++ b/deeptutor/services/config/provider_runtime.py@@ -380,11 +380,14 @@ STT_PROVIDERS: dict[str, VoiceProviderSpec] = {adapter="volcengine",default_model="bigmodel",),+ # DashScope's native adapter speaks the Paraformer real-time WebSocket API,+ # which only accepts `paraformer-realtime-*` ids. `paraformer-v2` is a batch+ # transcription id and the handshake fails with "Model not found"."dashscope": VoiceProviderSpec(label="Aliyun DashScope",default_api_base="https://dashscope.aliyuncs.com/api/v1",adapter="dashscope",- default_model="paraformer-v2",+ default_model="paraformer-realtime-v2",),"openai": VoiceProviderSpec(label="OpenAI",diff --git a/deeptutor/services/voice/options.py b/deeptutor/services/voice/options.pyindex f46de5fa1..29ddf29fd 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-v2"],+ "dashscope": ["paraformer-realtime-v2", "paraformer-realtime-8k-v2"],"volcengine_speech": ["bigmodel"],}.get(provider, [])languages = VOLC_LANGUAGES if provider == "volcengine_speech" else ISO_LANGUAGESdiff --git a/tests/api/test_settings_router.py b/tests/api/test_settings_router.pyindex a034819c4..2fec91d10 100644--- a/tests/api/test_settings_router.py+++ b/tests/api/test_settings_router.py@@ -648,7 +648,7 @@ def test_media_and_voice_provider_choices_include_dashscope() -> None:)assert dashscope["tts"]["default_model"] == "qwen3-tts-flash"assert dashscope["tts"]["default_voice"] == "Cherry"- assert dashscope["stt"]["default_model"] == "paraformer-v2"+ assert dashscope["stt"]["default_model"] == "paraformer-realtime-v2"assert dashscope["imagegen"]["default_model"] == "wanx2.1-t2i-turbo"assert dashscope["videogen"]["default_model"] == "wanx2.1-t2v-turbo"diff --git a/tests/services/test_voice.py b/tests/services/test_voice.pyindex 7f2c52f7e..115624c2c 100644--- a/tests/services/test_voice.py+++ b/tests/services/test_voice.py@@ -463,7 +463,7 @@ async def test_dashscope_stt_recognition_websocket_shape() -> None:websocket.send_str = record_start # type: ignore[method-assign]config = STTConfig(- model="paraformer-v2",+ model="paraformer-realtime-v2",provider_name="dashscope",adapter="dashscope",base_url="https://dashscope.aliyuncs.com/api/v1",@@ -474,7 +474,7 @@ async def test_dashscope_stt_recognition_websocket_shape() -> None:assert text == "hello world"start = json.loads(websocket.strings[0])- assert start["payload"]["model"] == "paraformer-v2"+ assert start["payload"]["model"] == "paraformer-realtime-v2"assert start["payload"]["parameters"] == {"format": "wav", "sample_rate": 16000}assert websocket.chunks == [b"RIFFxxxx"]assert json.loads(websocket.strings[-1])["header"]["action"] == "finish-task"@@ -578,7 +578,7 @@ def test_resolve_dashscope_voice_configs() -> None:"voice": "",}catalog["services"]["stt"]["profiles"][0]["binding"] = "bailian"- catalog["services"]["stt"]["profiles"][0]["models"][0]["model"] = "paraformer-v2"+ catalog["services"]["stt"]["profiles"][0]["models"][0]["model"] = "paraformer-realtime-v2"tts = resolve_tts_runtime_config(catalog=catalog)stt = resolve_stt_runtime_config(catalog=catalog)@@ -590,7 +590,7 @@ def test_resolve_dashscope_voice_configs() -> None:assert tts.base_url == "https://dashscope.aliyuncs.com/api/v1"assert stt.provider_name == "dashscope"assert stt.adapter == "dashscope"- assert stt.model == "paraformer-v2"+ assert stt.model == "paraformer-realtime-v2"assert stt.base_url == tts.base_url