--- title: "Create saved extension" method: POST path: "/v1/saved-extensions" tags: ["Digital Asset Management (DAM)", "Saved Extensions"] --- # Create saved extension `POST /v1/saved-extensions` This API creates a new saved extension. Saved extensions allow you to save complex extension configurations (like AI tasks) and reuse them by referencing the ID in upload or update file APIs. **Saved extension limit** \ You can create a maximum of 100 saved extensions per account. ## Request body - CreateSavedExtension - `name` string, required — Name of the saved extension. - `description` string, required — Description of what the saved extension does. - `config` union, required — Configuration object for an extension (base extensions only, not saved extension references). - object - `name` 'remove-bg', required — Specifies the background removal extension. - `options` object - `add_shadow` boolean — Whether to add an artificial shadow to the result. Default is false. Note: Adding shadows is currently only supported for car photos. - `semitransparency` boolean — Allows semi-transparent regions in the result. Default is true. Note: Semitransparency is currently only supported for car windows. - `bg_color` string — Specifies a solid color background using hex code (e.g., "81d4fa", "fff") or color name (e.g., "green"). If this parameter is set, `bg_image_url` must be empty. - `bg_image_url` string — Sets a background image from a URL. If this parameter is set, `bg_color` must be empty. - object - `name` 'google-auto-tagging' | 'aws-auto-tagging', required — Specifies the auto-tagging extension used. - `minConfidence` integer, required — Minimum confidence level for tags to be considered valid. - `maxTags` integer, required — Maximum number of tags to attach to the asset. - object - `name` 'ai-auto-description', required — Specifies the auto description extension. - object - `name` 'ai-tasks', required — Specifies the AI tasks extension for automated image analysis using AI models. - `tasks` union[], required — Array of task objects defining AI operations to perform on the asset. - union - AITaskSelectTags - `type` 'select_tags', required — Task type that analyzes the image and adds matching tags from a vocabulary. - `instruction` string, required — The question or instruction for the AI to analyze the image. - `vocabulary` string[] — Array of possible tag values. The combined length of all strings must not exceed 500 characters, and values cannot include the `%` character. When providing large vocabularies (more than 30 items), the AI may not follow the list strictly. - `min_selections` integer — Minimum number of tags to select from the vocabulary. - `max_selections` integer — Maximum number of tags to select from the vocabulary. - AITaskSelectMetadata - `type` 'select_metadata', required — Task type that analyzes the image and sets a custom metadata field value from a vocabulary. - `instruction` string, required — The question or instruction for the AI to analyze the image. - `field` string, required — Name of the custom metadata field to set. The field must exist in your account. - `vocabulary` union[] — An array of possible values matching the custom metadata field type. If not provided for SingleSelect or MultiSelect field types, all values from the custom metadata field definition will be used. When providing large vocabularies (above 30 items), the AI may not strictly adhere to the list. - … - `min_selections` integer — Minimum number of values to select from the vocabulary. - `max_selections` integer — Maximum number of values to select from the vocabulary. - AITaskYesNo - `type` 'yes_no', required — Task type that asks a yes/no question and executes actions based on the answer. - `instruction` string, required — The yes/no question for the AI to answer about the image. - `on_yes` AITaskAction — Defines actions to perform based on AI task results. - … - `on_no` AITaskAction — Defines actions to perform based on AI task results. - … - `on_unknown` AITaskAction — Defines actions to perform based on AI task results. - … ## Response `201` Saved extension created successfully. - SavedExtension — Saved extension object containing extension configuration. - `id` string — Unique identifier of the saved extension. - `name` string — Name of the saved extension. - `description` string — Description of the saved extension. - `config` union — Configuration object for an extension (base extensions only, not saved extension references). - object - `name` 'remove-bg', required — Specifies the background removal extension. - `options` object - `add_shadow` boolean — Whether to add an artificial shadow to the result. Default is false. Note: Adding shadows is currently only supported for car photos. - `semitransparency` boolean — Allows semi-transparent regions in the result. Default is true. Note: Semitransparency is currently only supported for car windows. - `bg_color` string — Specifies a solid color background using hex code (e.g., "81d4fa", "fff") or color name (e.g., "green"). If this parameter is set, `bg_image_url` must be empty. - `bg_image_url` string — Sets a background image from a URL. If this parameter is set, `bg_color` must be empty. - object - `name` 'google-auto-tagging' | 'aws-auto-tagging', required — Specifies the auto-tagging extension used. - `minConfidence` integer, required — Minimum confidence level for tags to be considered valid. - `maxTags` integer, required — Maximum number of tags to attach to the asset. - object - `name` 'ai-auto-description', required — Specifies the auto description extension. - object - `name` 'ai-tasks', required — Specifies the AI tasks extension for automated image analysis using AI models. - `tasks` union[], required — Array of task objects defining AI operations to perform on the asset. - union - AITaskSelectTags - `type` 'select_tags', required — Task type that analyzes the image and adds matching tags from a vocabulary. - `instruction` string, required — The question or instruction for the AI to analyze the image. - `vocabulary` string[] — Array of possible tag values. The combined length of all strings must not exceed 500 characters, and values cannot include the `%` character. When providing large vocabularies (more than 30 items), the AI may not follow the list strictly. - `min_selections` integer — Minimum number of tags to select from the vocabulary. - `max_selections` integer — Maximum number of tags to select from the vocabulary. - AITaskSelectMetadata - `type` 'select_metadata', required — Task type that analyzes the image and sets a custom metadata field value from a vocabulary. - `instruction` string, required — The question or instruction for the AI to analyze the image. - `field` string, required — Name of the custom metadata field to set. The field must exist in your account. - `vocabulary` union[] — An array of possible values matching the custom metadata field type. If not provided for SingleSelect or MultiSelect field types, all values from the custom metadata field definition will be used. When providing large vocabularies (above 30 items), the AI may not strictly adhere to the list. - … - `min_selections` integer — Minimum number of values to select from the vocabulary. - `max_selections` integer — Maximum number of values to select from the vocabulary. - AITaskYesNo - `type` 'yes_no', required — Task type that asks a yes/no question and executes actions based on the answer. - `instruction` string, required — The yes/no question for the AI to answer about the image. - `on_yes` AITaskAction — Defines actions to perform based on AI task results. - … - `on_no` AITaskAction — Defines actions to perform based on AI task results. - … - `on_unknown` AITaskAction — Defines actions to perform based on AI task results. - … - `createdAt` string, date-time — Timestamp when the saved extension was created. - `updatedAt` string, date-time — Timestamp when the saved extension was last updated. ## Other responses - `400` — Bad request. - `401` — Unauthorized request. - `403` — Forbidden. - `429` — The request exceeded the rate limit. Contains headers indicating the limits and a message detailing the error. --- [API](https://skmtc.dev/imagekit-developer/apis/imagekit-api.md) · [All operations](https://skmtc.dev/imagekit-developer/apis/imagekit-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/imagekit-developer/imagekit-api/revisions/bb3e9ff84b3e/schema)