add: openai speech to text

metagpt/provider/openai_api.py

@@ -239,3 +239,7 @@ class OpenAILLM(BaseLLM):
     async def atext_to_speech(self, **kwargs):
         """text to speech"""
         return await self.aclient.audio.speech.create(**kwargs)
+
+    async def aspeech_to_text(self, **kwargs):
+        """speech to text"""
+        return await self.aclient.audio.transcriptions.create(**kwargs)

tests/metagpt/provider/test_openai.py

@@ -1,5 +1,6 @@
 import pytest
 
+from metagpt.const import TEST_DATA_PATH
 from metagpt.llm import LLM
 from metagpt.logs import logger
 from metagpt.provider import OpenAILLM
@@ -53,6 +54,14 @@ async def test_text_to_speech():
     assert 200 == resp.response.status_code
 
 
+@pytest.mark.asyncio
+async def test_speech_to_text():
+    llm = LLM()
+    audio_file = open(f"{TEST_DATA_PATH}/audio/hello.mp3", "rb")
+    resp = await llm.aspeech_to_text(file=audio_file, model="whisper-1")
+    assert "你好" == resp.text
+
+
 class TestOpenAI:
     def test_make_client_kwargs_without_proxy(self):
         instance = OpenAILLM(mock_llm_config)