Defined in: packages/ai/src/utilities/embedding-input.ts:8
One embedding input item resolved into its text and image constituents. Produced by resolveEmbeddingInput; adapters map each entry onto one provider-native input (one vector per entry).
images: ImagePart<unknown>[];Defined in: packages/ai/src/utilities/embedding-input.ts:12
Image parts of the item, in order (empty for text-only items)
texts: string[];Defined in: packages/ai/src/utilities/embedding-input.ts:10
Text contents of the item, in order (empty for image-only items)