feat: add descs for speech to text responses

Author: TAnd-dev

packages/openapi/descriptions/speech-to-text.response.js

@@ -0,0 +1,41 @@
+export default {
+  metadata: {
+    desc: `Metadata about the transcription response, including timing, models, and IDs.`,
+  },
+  transaction_key: {
+    desc: `A unique transaction key; currently always “deprecated”.`,
+  },
+  request_id: {
+    desc: `A UUID identifying this specific transcription request.`,
+  },
+  sha256: {
+    desc: `The SHA-256 hash of the submitted audio file (for pre-recorded requests).`,
+  },
+  created: {
+    desc: `ISO-8601 timestamp.`,
+  },
+  duration: {
+    desc: `Length of the audio in seconds.`,
+  },
+  channels: {
+    desc: `Number of independent audio channels detected/submitted.`,
+  },
+  models: {
+    desc: `List of model UUIDs used for this transcription`,
+  },
+  model_info: {
+    desc: `Mapping from each model UUID (in 'models') to detailed info: its name, version, and architecture.`,
+    name: {
+      desc: `The human-readable name of the model — identifies which model was used.`,
+    },
+    version: {
+      desc: `The specific version of the model.`,
+    },
+    arch: {
+      desc: `The architecture of the model — describes the model family / generation.`,
+    },
+  },
+  status: {
+    desc: `The current status of the generation task.`,
+  },
+};

packages/openapi/index.js

@@ -16,4 +16,5 @@ export { default as videoResponse } from './descriptions/video.response.js';
 export { default as chatCompletionResponse } from './descriptions/chat-completion.response.js';
 export { default as imageResponse } from './descriptions/image.response.js';
 export { default as responsesResponse } from './descriptions/responses.response.js';
-export { default as audioResponse } from './descriptions/audio.response.js';
+export { default as audioResponse } from './descriptions/audio.response.js';
+export { default as speechToTextResponse } from './descriptions/speech-to-text.response.js';