In voice models and image generation models, the default request header issue with Omate makes it difficult to be compatible with other models.
For example, the URL for Silicon Flow’s voice model is “api.siliconflow.cn/v1/audio/speech”
Among them, the URL address entered in the software does not need to include /audio/speech
Omate’s request header to Silicon Flow automatically includes /audio/speech
However, if you want to switch to Volcano or Doubao’s model.
For example, Volcano’s request header is
“https://openspeech.bytedance.com/api/v3/tts/unidirectional”
No matter what you fill in the URL, it will automatically include /audio/speech, making the address request invalid.
Fangtang, can you adjust it here so that we can enter the complete address ourselves?