gradium

type STTConfig struct {
	// APIKey is the Gradium API key, sent as the x-api-key header. Required.
	APIKey string `validate:"required"`
	// URL overrides the ASR WebSocket endpoint; empty uses the hosted endpoint.
	URL string
	// Model is the transcription model; empty uses a default.
	Model string
	// Encoding is the base audio encoding ("pcm", "wav", or "opus"); empty uses
	// "pcm". For PCM the sample rate is appended to form the input format (for
	// example "pcm_16000").
	Encoding string
	// Language hints the spoken language; the zero value lets the server decide.
	// Verified codes are de, en, es, fr, and pt; other languages fall back to
	// the base code.
	Language language.Language
	// SampleRate is the input audio sample rate; 0 uses the transport's rate.
	SampleRate int
	// DelayInFrames delays text generation by this many 80 ms frames; nil uses a
	// default. Allowed values: 7, 8, 10, 12, 14, 16, 20, 24, 36, 48.
	DelayInFrames *int `validate:"omitempty,oneof=7 8 10 12 14 16 20 24 36 48"`
}

type TTSConfig struct {
	// APIKey is the Gradium API key, sent as the x-api-key header. Required.
	APIKey string `validate:"required"`
	// URL overrides the TTS WebSocket endpoint; empty uses the hosted endpoint.
	URL string
	// VoiceID is the voice identifier; empty uses a default voice.
	VoiceID string
	// JSONConfig is an optional JSON configuration string forwarded to the model;
	// empty omits it.
	JSONConfig string
}