Add support for `gemini-embedding-2-preview` which embeds text, images, and video into a unified vector space. Currently only text embeddings are supported via `gemini-embedding-001`. The new model uses `inline_data` parts for non-text inputs.
Add support for
gemini-embedding-2-previewwhich embeds text, images, and video into a unified vector space.Currently only text embeddings are supported via
gemini-embedding-001. The new model usesinline_dataparts for non-text inputs.