@effect-uai/google 0.3.0 → 0.5.0

package/src/models.ts

@@ -31,6 +31,90 @@ export type GoogleModel =
  *
  * Reference: https://ai.google.dev/gemini-api/docs/embeddings
  */
+/**
+ * Known Gemini TTS model identifiers. `(string & {})` tail keeps
+ * autocomplete on the literals while still accepting any string.
+ *
+ * Reference: https://ai.google.dev/gemini-api/docs/speech-generation
+ */
+export type GeminiTtsModel =
+  | "gemini-2.5-flash-preview-tts"
+  | "gemini-2.5-pro-preview-tts"
+  | "gemini-3.1-flash-tts-preview"
+  // eslint-disable-next-line @typescript-eslint/ban-types
+  | (string & {})
+
+/**
+ * Known Gemini models that accept audio input via `inlineData` /
+ * `fileData` on `generateContent`. Used by `GeminiTranscriber` —
+ * transcription is prompt-driven on top of these models, not a
+ * dedicated endpoint.
+ *
+ * Reference: https://ai.google.dev/gemini-api/docs/audio
+ */
+export type GeminiSttModel =
+  | "gemini-3-flash-preview"
+  | "gemini-3.1-pro-preview"
+  | "gemini-3.1-flash-lite-preview"
+  | "gemini-2.5-flash"
+  | "gemini-2.5-pro"
+  // eslint-disable-next-line @typescript-eslint/ban-types
+  | (string & {})
+
+/**
+ * 30 prebuilt voice names for Gemini TTS. No custom-voice / cloning path
+ * is exposed on this surface, so the tail is stock-only (no
+ * `(string & {})` escape).
+ *
+ * Reference: https://ai.google.dev/gemini-api/docs/speech-generation
+ */
+export type GeminiVoiceName =
+  | "Zephyr"
+  | "Puck"
+  | "Charon"
+  | "Kore"
+  | "Fenrir"
+  | "Leda"
+  | "Orus"
+  | "Aoede"
+  | "Callirrhoe"
+  | "Autonoe"
+  | "Enceladus"
+  | "Iapetus"
+  | "Umbriel"
+  | "Algieba"
+  | "Despina"
+  | "Erinome"
+  | "Algenib"
+  | "Rasalgethi"
+  | "Laomedeia"
+  | "Achernar"
+  | "Alnilam"
+  | "Schedar"
+  | "Gacrux"
+  | "Pulcherrima"
+  | "Achird"
+  | "Zubenelgenubi"
+  | "Vindemiatrix"
+  | "Sadachbia"
+  | "Sadaltager"
+  | "Sulafat"
+
+/**
+ * Known Lyria (music generation) model identifiers. `(string & {})` tail
+ * keeps autocomplete on the literals while still accepting any string.
+ *
+ * - `lyria-3-clip-preview` — fixed 30s output, MP3 only.
+ * - `lyria-3-pro-preview` — up to a couple of minutes, MP3 or WAV.
+ *
+ * Reference: https://ai.google.dev/gemini-api/docs/music-generation
+ */
+export type LyriaModel =
+  | "lyria-3-clip-preview"
+  | "lyria-3-pro-preview"
+  // eslint-disable-next-line @typescript-eslint/ban-types
+  | (string & {})
+
 export type GoogleEmbeddingModel =
   | "gemini-embedding-2"
   | "gemini-embedding-001"