--- title: "Create Image Generation" method: POST path: "/v1/flows/image" tags: ["Flows", "Image"] --- # Create Image Generation `POST /v1/flows/image` Start an image generation with the selected model. ## Headers - `xi-api-key` string, nullable — Your API key. This is required by most endpoints to access our API programmatically. You can view your xi-api-key using the 'Profile' tab on the website. ## Request body - union - GPTImage1Request — Request body for the OpenAI GPT Image 1 model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `images` ImageReference[] — Up to 5 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `mask` union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `quality` 'low' | 'medium' | 'high' — The quality of the output image. - `background` 'transparent' | 'opaque' | 'auto' — The background of the output image. With `auto`, the model picks the background that suits the image. - `aspect_ratio` '1:1' | '3:2' | '2:3' — The aspect ratio of the output image. - `model_id` 'gpt-image-1', required — The model to use for the generation. - GPTImage15Request — Request body for the OpenAI GPT Image 1.5 model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `images` ImageReference[] — Up to 5 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `mask` union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `quality` 'low' | 'medium' | 'high' — The quality of the output image. - `background` 'transparent' | 'opaque' | 'auto' — The background of the output image. With `auto`, the model picks the background that suits the image. - `aspect_ratio` '1:1' | '3:2' | '2:3' — The aspect ratio of the output image. - `model_id` 'gpt-image-1.5', required — The model to use for the generation. - GPTImage2Request — Request body for the OpenAI GPT Image 2 model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `images` ImageReference[] — Up to 10 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `mask` union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `quality` 'low' | 'medium' | 'high' — The quality of the output image. - `model_id` 'gpt-image-2', required — The model to use for the generation. - `aspect_ratio` 'auto' | '1:1' | '4:5' | '5:4' | '3:4' | '4:3' | '2:3' | '3:2' | '1:2' | '2:1' | '9:16' | '16:9' | '21:9' | '1:3' | '3:1' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `resolution` '1K' | '2K' | '4K' — The resolution of the output image. - Gemini25FlashImageRequest — Request body for the Google Gemini 2.5 Flash image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `model_id` 'gemini-2.5-flash-image', required — The model to use for the generation. - `images` ImageReference[] — Up to 5 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '2:3' | '3:2' | '3:4' | '4:3' | '4:5' | '5:4' | '9:16' | '16:9' | '21:9' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - Gemini3ProImageRequest — Request body for the Google Gemini 3 Pro image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `model_id` 'gemini-3-pro-image', required — The model to use for the generation. - `images` ImageReference[] — Up to 10 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '2:3' | '3:2' | '3:4' | '4:3' | '4:5' | '5:4' | '9:16' | '16:9' | '21:9' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `resolution` '1K' | '2K' | '4K' — The resolution of the output image. - Gemini31FlashImageRequest — Request body for the Google Gemini 3.1 Flash image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `model_id` 'gemini-3.1-flash-image', required — The model to use for the generation. - `images` ImageReference[] — Up to 14 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '2:3' | '3:2' | '3:4' | '4:3' | '4:5' | '5:4' | '9:16' | '16:9' | '21:9' | '1:4' | '4:1' | '1:8' | '8:1' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `resolution` '512' | '1K' | '2K' | '4K' — The resolution of the output image. - Gemini31FlashLiteImageRequest — Request body for the Google Gemini 3.1 Flash Lite image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `model_id` 'gemini-3.1-flash-lite-image', required — The model to use for the generation. - `images` ImageReference[] — Up to 14 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '2:3' | '3:2' | '3:4' | '4:3' | '4:5' | '5:4' | '9:16' | '16:9' | '21:9' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `resolution` '1K' — The resolution of the output image. - BytedanceSeedream5LiteRequest — Request body for the ByteDance Seedream 5.0 Lite image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `images` ImageReference[] — Up to 10 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '3:4' | '16:9' | '4:3' | '9:16' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `seed` integer, nullable — A seed for reproducible generation: the same seed and inputs give similar output across generations. Omit for random. - `model_id` 'bytedance-seedream-5-lite', required — The model to use for the generation. - `resolution` '2K' | '3K' — The resolution of the output image. - BytedanceSeedream5ProRequest — Request body for the ByteDance Seedream 5.0 Pro image model. - `webhook` union - WebhookTargetAll — Deliver the result to all of the workspace's configured flows webhooks. - `type` 'all' — Send the result to all of the workspace's configured flows webhooks. - WebhookTargetIds — Deliver the result to specific configured flows webhooks. - `type` 'ids' — Send the result to the listed flows webhooks. - `ids` string[], required — The IDs of the workspace flows webhooks to deliver the result to. Each must be one of the workspace's configured flows webhooks. - `prompt` string, required — A text description of the image to generate. - `images` ImageReference[] — Up to 10 reference images to edit or draw from. - union - GenerationReference — The output of a prior generation on this API. - `type` 'generation', required - `generation_id` string, required — The ID of the generation whose output to use, as returned when the generation was created. - StaticAssetReference — An asset uploaded via the assets API. - `type` 'asset', required - `asset_id` string, required — The ID of an asset uploaded via the assets API (`POST /v1/assets`), as returned in that response's `asset_id`. - InlineImageReference — An image passed inline as base64. The image is stored as an ephemeral asset with no guaranteed retention: it may be deleted at any time after the generation completes. To keep an input and reuse it across generations, upload it via the assets API (`POST /v1/assets`) and pass an `asset` reference instead. - `type` 'inline_base64', required - `content_base64` string, required — The media file's bytes, base64-encoded (standard alphabet). Up to 25MB decoded. - `mime_type` 'image/jpeg' | 'image/png' | 'image/webp' | 'image/heic' | 'image/heif', required — The MIME type of the encoded image. - `aspect_ratio` 'auto' | '1:1' | '3:4' | '16:9' | '4:3' | '9:16' — The aspect ratio of the output image. With `auto`, the model picks an aspect ratio based on the inputs. - `seed` integer, nullable — A seed for reproducible generation: the same seed and inputs give similar output across generations. Omit for random. - `model_id` 'bytedance-seedream-5-pro', required — The model to use for the generation. - `resolution` '1K' | '2K' — The resolution of the output image. ## Response `200` Successful Response - MediaGenerationCreateResponse — A newly queued media generation; fetch the GET endpoint for the output. - `id` string, required — The unique identifier of the generation. Pass it to the corresponding GET endpoint to retrieve the output. - `status` 'pending', required — A newly created generation is always `pending`. ## Other responses - `422` — Validation Error --- [API](https://skmtc.net/elevenlabs/apis/elevenlabs-api-documentation.md) · [All operations](https://skmtc.net/elevenlabs/apis/elevenlabs-api-documentation/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/elevenlabs/elevenlabs-api-documentation/revisions/acb0175c0525/schema)