> ## Documentation Index > > Fetch the complete documentation index at: https://kling.ai/document-api/llms.txt > Use this file to discover all available pages before exploring further. # Image to Video > Source: https://kling.ai/document-api/api/video/4-0/image-to-video > Locale: en > Current Tab: Image to Video > Sibling Tabs: Text to Video / Image to Video / Omni Video Generation / Element Management / Voice Management > This content is optimized for LLMs. In-page tabs are expanded and UI-only controls are omitted. --- ## Create Task ### API Overview - Method: `POST` - Path: `/image-to-video/kling-4.0` - Auth: `Authorization: Bearer ` - Content-Type: `application/json` ### Description Supports video generation from multiple keyframes, including first frame, middle frame and last frame, with up to 10 images per request. ### Headers | Field | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `Content-Type` | string | Yes | `application/json` | - | Data Exchange Format | | `Authorization` | string | Yes | - | - | Authentication information, refer to API authentication | ### Request Body | Field Path | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `contents` | array | Yes | - | - | Reference input collection, such as prompt and images | | `contents[].type` | string | Yes | - | `prompt`, `first_frame`, `middle_frame`, `last_frame` | Input type. Supports: prompt, first frame, middle frame and last frame | | `settings` | object | No | - | - | Output configuration-related parameters, such as resolution, duration, etc | | `settings.resolution` | string | No | `1080p` | `720p`, `1080p` | The resolution of the generated video | | `settings.duration` | int | No | `10` | `3`, `4`, `5`, `6`, `7`, `8`, `9`, `10`, `11`, `12`, `13`, `14`, `15`, `16`, `17`, `18`, `19`, `20`, `21`, `22`, `23`, `24`, `25`, `26`, `27`, `28`, `29`, `30` | Video length, unit: s (seconds) | | `options` | object | No | - | - | General configurations, such as callback address, whether to include watermark, etc | | `options.callback_url` | string | No | - | - | Callback notification URL for task result | | `options.external_task_id` | string | No | - | - | Customized Task ID | | `options.watermark_info` | object | No | - | - | Whether to generate watermarked results simultaneously | | `options.watermark_info.enabled` | boolean | No | `false` | - | true: generate watermarked result; false: do not generate | #### Request Body Field Notes - `contents`: The reference format is as follows, with detailed parameter descriptions provided below: - `contents`: ```json "contents": [ { "type": "prompt", "text": "string" }, { "type": "first_frame", "url": "string", "id": "string" }, { "type": "middle_frame", "url": "string", "id": "string" }, { "type": "last_frame", "url": "string", "id": "string" } ] ``` - `contents[].type.prompt`: English-only input supports up to 26,000 characters, and non-English-only input supports up to 11,000 characters. Mixed input must satisfy E / 26000 + N / 11000 <= 1, where E is the number of English characters and N is the number of non-English characters. - `contents[].type.prompt`: Defined in JSON format, as follows: - `contents[].type.prompt`: ```json { "type": "prompt", // Input type, fixed parameter value: prompt; required "text": "string" // Text prompt content; required } ``` - `contents[].type.prompt`: > **The Kling 4.0 API supports multi-reference video generation and video editing through prompts that describe reference roles, scene content, event sequences, and audio requirements.** > > > 1. **Reference input materials using custom IDs.** When defining an image, video, or element object in `contents`, assign it an `id` and use `@id` to reference it in the prompt. For example, set `"id": "image_1"` on a `refer_image` object, then reference it as `@image_1`. IDs must be unique within the same task. > > > 2. **Specify the purpose and scope of each reference.** For example: “Preserve the person’s facial features, hairstyle, and clothing from `@image_1`” or “Follow the movement rhythm in `@video_1` without using its background.” If a material is referenced without an explicit purpose, the model will infer how to use it, which may produce results that differ from your intent. > > > 3. **Organize prompts around reference roles, overall settings, event sequences, and audio requirements.** Describe the scene, visual style, camera movement, action sequence, dialogue, ambient sounds, music, and sound effects as needed. For video editing, explicitly specify what to change and what to preserve. For more details, see the [Kling 4.0 User Guide](https://kling.ai/quickstart/kling-4-model-user-guide). - `contents[].type.first_frame / middle_frame / last_frame`: Defined in JSON format, as follows: - `contents[].type.first_frame / middle_frame / last_frame`: ```json { "type": "string", // Frame image identifier. Must match enum values: first_frame, middle_frame, last_frame; required "url": "string", // The content of the input can be provided via URL or Base64; required "id": "string" // Reference ID for this input, used to reference it in the prompt (for example, @image_1); must be unique within the task; optional } ``` - `contents[].type.first_frame / middle_frame / last_frame`: Image format: .jpg, .jpeg, .png. - `contents[].type.first_frame / middle_frame / last_frame`: The size of the image file cannot exceed 30MB. - `contents[].type.first_frame / middle_frame / last_frame`: The width and height dimensions of the image should not be less than 300px, and the aspect ratio of the image should be between 2:5 and 5:2. - `contents[].type.first_frame / middle_frame / last_frame`: Maximum number of images: 10. - `settings.resolution`: 720p: output video with a resolution of 720P - `settings.resolution`: 1080p: output video with a resolution of 1080P - `settings.duration`: When this parameter is provided, the specified duration takes precedence. - `settings.duration`: When this parameter is omitted, the model detects and follows the duration intent in the Prompt, for example, "Generate a 15-second video." - `settings.duration`: If no explicit duration intent is detected in the Prompt, the default duration of 10 seconds is used. - `options.callback_url`: If configured, the server will actively notify when the task status changes - `options.callback_url`: For specific message schema, see [Callback Protocol](https://kling.ai/document-api/api/get-started/callbacks) - `options.external_task_id`: Will not overwrite the system-generated task ID, but can be used for task queries - `options.external_task_id`: Must be unique within a single user account - `options.watermark_info`: Defined through the enabled parameter: ```json "watermark_info": { "enabled": false // true: generate watermarked result; false: do not generate; default: false } ``` - `options.watermark_info`: Custom watermarks are not currently supported ### Request Example ```bash curl --location 'https://api-singapore.klingai.com/image-to-video/kling-4.0' \ --header 'Content-Type: application/json' \ --header "Authorization: Bearer ${KLING_API_KEY}" \ --data '{ "contents": [ { "type": "prompt", "text": "A girl sat on the train, looking out the window with a melancholic expression, her head swaying with the train." }, { "type": "first_frame", "url": "https://p2-kling.klingai.com/kcdn/cdn-kcdn112452/kling-tob-release_note/image_25.png" } ], "settings": { "resolution": "1080p", "duration": 10 }, "options": { "callback_url": "https://xxx/callback", "external_task_id": "", "watermark_info": { "enabled": false } } }' ``` ### Response Example ```json { "code": 0, // Error code. See Error Codes for details "message": "string", // Error information "request_id": "string", // Request ID, generated by the system "data": { "id": "string", // The task ID generated by the system "status": "string", // Task status, Enum values: submitted, processing, succeeded, failed "create_time": 1781080778802, // Task creation time, Unix timestamp, unit ms "update_time": 1781080794151, // Task update time, Unix timestamp, unit ms "external_id": "string" // The custom task ID for this task (if any) } } ``` ## More Scenario Examples ### Multi-Keyframes: First + Middle + Last Frame ```cURL curl --location 'https://api-singapore.klingai.com/image-to-video/kling-4.0' \ --header "Authorization: Bearer ${KLING_API_KEY}" \ --header 'Content-Type: application/json' \ --data-raw '{ "contents": [ { "type": "prompt", "text": "A cinematic, photorealistic fashion video. Begin with the female model in @image_1, preserving her facial features, hairstyle, clothing, and accessories. She moves naturally and gracefully, transitioning smoothly into the pose and composition shown in @image_2. Continue the motion toward the final pose and scene shown in @image_3. Ensure seamless transitions between all three keyframes, with realistic body movements, natural facial expressions, consistent character identity, and coherent lighting. Smooth camera movement, premium fashion commercial aesthetic. Avoid abrupt cuts, distorted anatomy, and unnatural motion." }, { "type": "first_frame", "url": "https://p2-kling.klingai.com/kcdn/cdn-kcdn112452/kling-qa-test/file_size/person-ref-pad-2x3-2MB.jpg", "id": "image_1" }, { "type": "middle_frame", "url": "https://p4-kling.klingai.com/kcdn/cdn-kcdn112452/chrome-plugin-upload/2026-10-13/1791892036356.caca59ef0c55e171.png", "id": "image_2" }, { "type": "last_frame", "url": "https://p2-kling.klingai.com/kcdn/cdn-kcdn112452/chrome-plugin-upload/2026-10-13/1791892087071.f32f7a727b1832bf.png", "id": "image_3" } ], "options": { "external_task_id": "i2v_v4_three_frames_1013", "watermark_info": { "enabled": false } } }' ``` --- ## Query Task (By task ID) ### API Overview - Method: `GET` - Path: `/tasks` - Auth: `Authorization: Bearer ` - Content-Type: `application/json` ### Description > Note: The API currently supports querying async tasks only ### Headers | Field | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `Content-Type` | string | Yes | `application/json` | - | Data Exchange Format | | `Authorization` | string | Yes | - | - | Authentication information, refer to API authentication | ### Query Params | Field | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `task_ids` | string | No | - | - | The task ID defined by the system that needs to be queried | | `external_task_ids` | string | No | - | - | The custom task ID that needs to be queried | #### Query Params Field Notes - `task_ids`: Query parameter; provide the value in the URL query string - `task_ids`: When querying a task, you can choose between two query methods: task_ids and external_task_ids, but you cannot use them simultaneously - `task_ids`: Support batch queries, separated by "," - `external_task_ids`: Query parameter; provide the value in the URL query string - `external_task_ids`: When querying a task, you can choose between two query methods: task_ids and external_task_ids, but you cannot use them simultaneously - `external_task_ids`: Support batch queries, separated by "," ### Request Example ```bash curl --location 'https://api-singapore.klingai.com/tasks?external_task_ids=123' \ --header 'Content-Type: application/json' \ --header "Authorization: Bearer ${KLING_API_KEY}" ``` ### Response Example ```json { "code": 0, // Error code. See Error Codes for details "message": "string", // Error information "request_id": "string", // Request ID, generated by the system, is used to track requests and troubleshoot problems "data": [ { "id": "893605946402811985", // The task ID being queried "status": "string", // Task status, Enum values: submitted, processing, succeeded, failed "message": "string", // Task status information, displaying the failure reason when the task fails (such as triggering the content risk control of the platform, etc.) "create_time": 1781080778802, // Task creation time, Unix timestamp, unit ms "update_time": 1781080794151, // Task update time, Unix timestamp, unit ms "external_id": "string", // The custom task ID for this task (if any) "outputs": [ { "type": "video", // When the generated result is "video", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Video ID, generated by the system "url": "string", // Generated result URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "watermark_url": "string", // URL for generating result with watermark, hotlink protection format "duration": "string" // The duration of the generated video, in seconds }, { "type": "image", // When the generated result is "image", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "url": "string", // Generated result URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "watermark_url": "string", // Image download URL with watermark, protected URL "group_id": "string" // Only appears when generating grouped images, used to mark grouping relationships }, { "type": "audio", // When the generated result is "audio", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Audio ID, generated by the system "mp3_url": "string", // Generated MP3 URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "wav_url": "string", // Generated WAV URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "mp3_duration": "string", // The duration of the generated MP3 format audio, in seconds "wav_duration": "string" // The duration of the generated WAV format audio, in seconds }, { "type": "voice", // When the generated result is "voice", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Voice ID, generated by the system "name": "string", // Voice name "url": "string", // Material download link "owned_by": "string", // Voice source, kling is the official voice library, and the number is the creator ID "status": "succeeded" // Voice status. Enum values: succeeded, deleted }, { "type": "element", // When the generated result is "element", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Element ID, generated by the system "name": "string", // Element name "description": "string", // Element description "element_type": "string", // Element type, divided into video character element and multi-image element, enumeration values are: video_character_elements and multi_image_elements "references": [ // Related materials of element { "type": "image", // When returning 'image' materials, the enumeration values for each content type are: image, video, voice "role": "string", // Image reference material attributes are divided into frontal reference images and other reference images, with enumeration values of frontal and reference, respectively "url": "string" // Material download link }, { "type": "video", // When returning "video" materials, the enumeration values for each content type are: image, video, voice "role": "refer", // Video reference material attribute, fixed value: refer "url": "string" // Material download link }, { "type": "voice", // When "voice" material is returned, enumeration values of each content type: image, video, voice "role": "refer", // Voice reference material attribute, fixed value: refer "url": "string", // Material download link "id": "string", // Voice ID "name": "string", // Voice name "owned_by": "string" // Voice source. kling means official voice library; numbers are creator IDs } ], "owned_by": "string", // Element source. kling means the official Element library; numbers are creator IDs "status": "string", // Element status. Enum values: succeeded, deleted "tags": [ // Element label related information { "id": 1, // Tag ID "name": "string", // Tag name "description": "string" // Tag description } ] } ], "billing": [ { "charge_type": "string", // Consumption account type: If the consumption is of balance, the parameter value is "cash"; if it is of resource package, the parameter value is "unit" "cash_type": "string", // Balance type, only exists in the consumption limit scenario (charge_type=cash). If the consumption is the official quota, the parameter value is balance. If the consumption is the test quota, the parameter value is test_balance "amount": "string", // Deduction amount; In the balance deduction situation (charge_type=cash), it represents the discount price deducted from the balance, and in the resource package deduction situation (charge_type=unit), it represents the deduction amount of units; decimal system "currency": "string", // Consumption unit, which only exists in the consumption balance scenario (charge_type=cash), with a fixed parameter value of CNY/USD "package_type": "string", // Consumable resource bundle type, which only exists in the consumable resource bundle scenario (charge_type=unit), with fixed enumeration values: video, image, audio "list_price": "string" // Balance deduction list price only exists in the consumption limit scenario (charge_type=cash) } ] } ] } ``` --- ## Query Task (By Cursor) ### API Overview - Method: `POST` - Path: `/tasks` - Auth: `Authorization: Bearer ` - Content-Type: `application/json` ### Description > Note: The API currently supports querying async tasks only ### Headers | Field | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `Content-Type` | string | Yes | `application/json` | - | Data Exchange Format | | `Authorization` | string | Yes | - | - | Authentication information, refer to API authentication | ### Request Body | Field Path | Type | Required | Default | Enum | Description | |---|---|---:|---|---|---| | `start_time` | string | No | `end_time - 30 days` | - | Start time for task creation filtering | | `end_time` | string | No | `current time` | - | End time for task creation filtering | | `cursor` | string | No | - | - | Continuation cursor, i.e. query starting point | | `limit` | int | No | `100` | - | Query the number of tasks | | `filters` | array | No | - | - | Query task filtering criteria, such as task status and function type | | `filters[].key` | string | No | - | `status`, `product_type` | Filtering dimensions, currently supporting filtering based on task status and other conditions | | `filters[].values` | array | No | - | - | Filter value | #### Request Body Field Notes - `start_time`: Unix timestamp, in milliseconds - `start_time`: The default value is end_time - 30 days - `start_time`: Start time must be earlier than end time - `end_time`: The default value is current time - `end_time`: Unix timestamp, in milliseconds - `end_time`: End time must be later than start time - `cursor`: Use the next_cursor returned by the previous query - `cursor`: When cursor is not empty, start_time and end_time are ignored - `limit`: Maximum value: 500 - `filters`: Load with key/value format: ```json "filters": [ { "key": "status", // Filter dimension by task status, fixed value: status "values": ["succeeded"] }, { "key": "product_type", // Filter dimension by function type, fixed value: product_type "values": ["video"] } ] ``` - `filters[].key`: status: filter by task status - product_type: filter by function type - `filters[].values`: status: submitted, processing, succeeded, failed - `filters[].values`: product_type: video, image, try_on ### Request Example ```bash curl --location 'https://api-singapore.klingai.com/tasks' \ --header 'Content-Type: application/json' \ --header "Authorization: Bearer ${KLING_API_KEY}" \ --data '{ "start_time": "1781193600000", "end_time": "1781516352968", "cursor": "", "limit": 500, "filters": [ { "key": "status", "values": ["succeeded"] }, { "key": "product_type", "values": ["video"] } ] }' ``` ### Response Example ```json { "code": 0, // Error code. See Error Codes for details "message": "string", // Error information "request_id": "string", // Request ID, generated by the system "data": { "result": [ { "id": "string", // The task ID being queried "status": "string", // Task status, Enum values: submitted, processing, succeeded, failed "message": "string", // Task status information, displaying the failure reason when the task fails "create_time": 1781080778802, // Task creation time, Unix timestamp, unit ms "update_time": 1781080794151, // Task update time, Unix timestamp, unit ms "external_id": "string", // The custom task ID for this task (if any) "outputs": [ { "type": "video", // When the generated result is "video", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Video ID, generated by the system "url": "string", // Generated result URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "watermark_url": "string", // URL for generating result with watermark, hotlink protection format "duration": "string" // The duration of the generated video, in seconds }, { "type": "image", // When the generated result is "image", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "url": "string", // Generated result URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "watermark_url": "string", // Image download URL with watermark, protected URL "group_id": "string" // Only appears when generating grouped images, used to mark grouping relationships }, { "type": "audio", // When the generated result is "audio", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Audio ID, generated by the system "mp3_url": "string", // Generated MP3 URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "wav_url": "string", // Generated WAV URL with hotlink protection. (Generated results are retained for 30 days; save them promptly.) "mp3_duration": "string", // The duration of the generated MP3 format audio, in seconds "wav_duration": "string" // The duration of the generated WAV format audio, in seconds }, { "type": "voice", // When the generated result is "voice", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Voice ID, generated by the system "name": "string", // Voice name "url": "string", // Material download link "owned_by": "string", // Voice source, kling is the official voice library, and the number is the creator ID "status": "succeeded" // Voice status. Enum values: succeeded, deleted }, { "type": "element", // When the generated result is "element", the return value and related fields vary by content type. Enum values: image, video, audio, element, voice "id": "string", // Element ID, generated by the system "name": "string", // Element name "description": "string", // Element description "element_type": "string", // Element type, divided into video character element and multi-image element, enumeration values are: video_character_elements and multi_image_elements "references": [ // Related materials of element { "type": "image", // When returning 'image' materials, the enumeration values for each content type are: image, video, voice "role": "string", // Image reference material attributes are divided into frontal reference images and other reference images, with enumeration values of frontal and reference, respectively "url": "string" // Material download link }, { "type": "video", // When returning "video" materials, the enumeration values for each content type are: image, video, voice "role": "refer", // Video reference material attribute, fixed value: refer "url": "string" // Material download link }, { "type": "voice", // When "voice" material is returned, enumeration values of each content type: image, video, voice "role": "refer", // Voice reference material attribute, fixed value: refer "url": "string", // Material download link "id": "string", // Voice ID "name": "string", // Voice name "owned_by": "string" // Voice source. kling means official voice library; numbers are creator IDs } ], "owned_by": "string", // Element source. kling means the official Element library; numbers are creator IDs "status": "string", // Element status. Enum values: succeeded, deleted "tags": [ // Element label related information { "id": 1, // Tag ID "name": "string", // Tag name "description": "string" // Tag description } ] } ], "billing": [ { "charge_type": "string", // Consumption account type: cash for balance, unit for resource package "cash_type": "string", // Balance type, only exists in the consumption limit scenario (charge_type=cash). If the consumption is the official quota, the parameter value is balance. If the consumption is the test quota, the parameter value is test_balance "amount": "string", // Deduction amount; In the balance deduction situation (charge_type=cash), it represents the discount price deducted from the balance, and in the resource package deduction situation (charge_type=unit), it represents the deduction amount of units; decimal system "currency": "string", // Consumption unit, which only exists in the consumption balance scenario (charge_type=cash), with a fixed parameter value of CNY/USD "package_type": "string", // Consumable resource bundle type, which only exists in the consumable resource bundle scenario (charge_type=unit), with fixed enumeration values: video, image, audio "list_price": "string" // Balance deduction list price only exists in the consumption limit scenario (charge_type=cash) } ] } ], "count": 1, // Number of query results "next_cursor": "string", // Cursor for querying subsequent results "has_more": true // Whether there are more results based on cursor information } } ```