:::tip tip **Viking AI Search** is an enterprise\-level AI search product powered by generative AI. It leverages the multimodal semantic understanding of large models and ByteDance's extensive experience in information retrieval to offer an out\-of\-the\-box platform. This service allows customers to easily build and integrate end\-to\-end multimodal search and recommendation capabilities for both images and text. * To explore more features of Viking AI Search, visit: [Viking AI Search Engine, the all-in-one intelligent engine for enterprise search, recommendations, and Q&A](https://www.volcengine.com/product/AI-Search-Rec) ::: Searches for content related to the user's input query or image within the item data source bound to the application. ## Request description * Request method: POST * Request URL: * North China: https://aisearch.cn\-beijing.volces.com/api/v1/application/`{application_id}`/search/{scene_id} * Johor: https://aisearch.ap\-southeast\-1.volces.com/api/v1/application/`{application_id}`/search/{scene_id} `{application_id}:` the unique identifier for an application, assigned by the system after creation. You can view this ID in the console. `{scene_id}:` the unique policy identifier assigned by the system after a new search policy is created on the experience configuration page in the console. You can find it by navigating to "Search Experience" in the console, hovering over the policy name, and copying the policy ID. ![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/0d5092219f3f4598a24c580a41e33a95 =2896x) ## Notes * **Scope of Use:** * Only available for Item datasets. Search is not supported for Document and User Behavior datasets. * **Request Body Limit:** * The body of a single request cannot exceed 10 MB. ## API Rate Limiting The call rate limit is 20 QPS per tenant. Calls that exceed the API rate limit will return a 429 status code. ## Authentication Method Use an API Key or Access Key for authentication. For details, refer to the [API Authentication Guide](https://www.volcengine.com/docs/85296/1544950). ## Request Parameters ### Path Parameters |**Parameter** |**Type** |**Required** |**Example** |**Description** | |---|---|---|---|---| |`application_id` |String |%%require%% |76414365027 |Your unique application identifier. | ### Body Parameters |**Parameter** |**Type** |**Required** |**Example value** |**Description** | |---|---|---|---|---| |`query` |[Query](/docs/85296/1544974#7d914235) |%%require%% | |The text or image information for the search query passed in the Query object. Options:|\ | | | |```Plain Text||\ | | | |{|* Text search only|\ | | | | "text":"同款鞋",|* Image search only|\ | | | | "image_url":{base64图片文件}|* Image and text search|\ | | | |}||\ | | | |```|For parameter details, see: [Query](/docs/85296/1544974#7d914235) |\ | | | | | | |`dataset_id` |string |%%require%% |"106265704" |The dataset ID can be viewed in the dataset list or on the dataset details page. When using the direct search feature, you must explicitly specify the dataset ID in each request. | |`page_number` |int |%%require%% |1 |The page number for search results, used for list pagination, starting from 1. For example: "page_number": 1 returns the first page of search results. | |`page_size` |int |%%require%% |20 |The number of search results per page. You can enter a value from 1 to the maximum number of results you set in the console.|\ | | | | |To ensure optimal performance and reduce loading times on the client side, we recommend setting the page size to a maximum of 100 and loading additional data through pagination.|\ | | | | |Default: 10 | |`user` |[User](/docs/85296/1544974#9dc80774) |Optional |```JSON|Enter the User object with the searching user's information. You can upload the user's pseudo ID to specify which user triggered this API call. |\ | | | |{| |\ | | | | "_user_id":"U90E89FF8045"| |\ | | | |}| |\ | | | |```| |\ | | | | | | |`filter` |[Filter](/docs/85296/1544974#aead416b) |Optional |See the data structure below\-[Filter](/docs/85296/1544974#aead416b) |A search filter object that filters search results based on the filterable fields configured in the application's data source.|\ | | | | ||\ | | | | |* The default value is empty, meaning no filter is applied to the search results.|\ | | | | |* Filter conditions support the `must`, `must_not`, and `range` query operators, as well as the `and` and `or` logical operators. You can combine these query operators. | |`sort_by` |string |Optional |"create_at" |The name of the field used for sorting. **The field must exist in the dataset and have "Used for filtering" selected in the application.** The following types are supported:|\ | | | | ||\ | | | | |* **Basic types**: `int32`, `int64`, `float`, `bool`, `string`|\ | | | | |* **Complex types**: To sort the sub\-fields of an `object` type, use a `.` The connector references internal fields using the format `object.field`, for example, `product.price`. | |`sort_order` |string |Optional |"desc" |Sort order. Supported enumerated values: "desc"\|"asc":|\ | | | | ||\ | | | | |* "desc": Descending order (default if not specified)|\ | | | | |* "asc": Ascending order | |`output_fields` |array(string) |Optional |["title", "price", "category"]|Specify which fields from the item dataset to return for display in 'display_fields'. If this parameter is not provided, all fields will be returned.|\ | | | | |> If the dataset contains nested objects (object or array), you can only pass top\-level field names.|\ | | | | |> Fields that do not exist or are nested within an object will be ignored. | |`context` |object(context) |Optional |See the data structure below: [Context](/docs/85296/1544974#7f9aa6e1) |Configures the request context. It currently supports passing the geographic location of the user initiating the search. | |`conditional_boost` |array(conds) |Optional |See: [Conds](/docs/85296/2166627#ca81d523) |Request\-level promotion and demotion configuration|\ | | | | |> Learn more: [AI Search: Boost and Demotion](/docs/85296/2166627)|\ | | | | ||\ | | | | |The platform automatically merges promotion and demotion rule configurations from the console and API, and they take effect simultaneously. | |`disable_personalize` |Boolean |Optional |false |Enable/Disable the **personalized** search capability.|\ | | | | | Default is **false**.|\ | | | | ||\ | | | | |* When the parameter is **true**, search results will be ranked based only on: |\ | | | | | * **Item relevance** (default logic, cannot be disabled) |\ | | | | | * **Item popularity** (only if this feature has been enabled in the console; otherwise it will not take effect. See [Sorting based on Item Popularity](/docs/85296/1873485#ef8c5c0a)) |\ | | | | ||\ | | | | | |\ | | | | | |\ | | | | |* When the parameter is **false**, in addition to the above, search results can also be influenced by: |\ | | | | | * **Personalized retrieval** (only if this feature has been enabled in the console; otherwise it will not take effect. See [Configure Personalized Retrieval](/docs/85296/1873485#4b9c35f1)) | ### Data Structure #### Query |**Parameter** |**Type** |**Required** |**Example value** |**Description** | |---|---|---|---|---| |text |string |At least one is required |"Women's shoes spring/summer" "The Three\-Body Problem" |Text query | |image_url |string |^^|"data:image/jpeg;base64,iVBUAAAA..."|Image Search|\ | | | |"www.example.com/image.png" |**Applications linked to video datasets do not support ** **image search.** | |image_query_instruction |string |Optional when the `image_url` field is provided |"Retrieve similar images based on their overall content, rather than focusing on specific details within the image" |An instruction for image search used to enhance search performance in search\-by\-image scenarios | ##### **Uploading Images in Query** For the image passed in `query.image_url`, we support either an **image URL (recommended)** or a **base64\-encoded** image file. :::tip Limits **Input images must meet the following conditions** **:** **Supported formats:** jpeg, png, webp, bmp, tiff, icns. For tiff and icns images, ensure they align with their metadata by correctly setting the file metadata in object storage. Otherwise, parsing will fail. **Aspect ratio (width/height):** Within the range [1/100, 100]. **Side length:** [10, 6000] px. The shorter side must be greater than 10 px, and the longer side must be less than 6000 px. ::: **Option 1: Image URL (Recommended)** Ensure the image URL is publicly accessible and triggers a preview instead of a download when accessed. **Option 2: Base64 Format** Ensure the Base64 encoding is in the standard format, including two parts: **a data header (Data URI Scheme)** and **a Base64\-encoded body**. ```Plain Text data:[];base64, ``` |**Components** |**Format requirements** |**Description** |**Example** | |---|---|---|---| |Data header prefix |Must be `data:` |Indicates the string is in Data URI format and cannot be omitted or modified |\- | |MIME Type |Must strictly match the actual image format. |Used by the server to identify the image type. Do not enter incorrect information, such as using the PNG MIME type for a JPG image. |JPG/JPEG：`image/jpeg`；PNG：`image/png`；WebP：`image/webp` | |Separator |Always `;base64,` |Connects the MIME type with the Base64 encoded body. The semicolon, "base64", and comma cannot be omitted or replaced. |\- | |Base64 Encoded Body |A Base64\-encoded string that complies with the RFC 4648 standard. |Generated by directly encoding the binary data of an image file. **It cannot contain spaces, line breaks, or special characters** and must be a continuous ASCII string. | | #### User |**Parameter** |**Type** |**Required** |**Example value** |**Description** | |---|---|---|---|---| |`_user_id` |string |Optional |"ldy_199432" |This represents **a unique identifier for the end\-user**.|\ | | | | |This field specifies the user who triggered the API call, which is crucial for distinguishing between different users from the same organization.|\ | | | | |For example, it can help you record detailed operation logs, provide personalized recommendations, or analyze different user behavior patterns. We recommend assigning and passing a unique ID for each end\-user. | #### Filter |**Parameter** |**Type** |**Required** |**Example value** |**Description** | |---|---|---|---|---| |`op` |string |%%require%% |"must" |The supported operator types are:|\ | | | | |"must"、"must_not"、"range"、"and"、"or" | |field |string |Conditionally required |"status" |The field to which the filter condition is applied.|\ | | | | ||\ | | | | |* Required for the "must", "must_not", and "range" operators.|\ | | | | |* The "and" and "or" operators ignore this parameter. | |gte |Float |Optional |100.0 |The filter range for the "range" operator, which filters data where the "field" value is `greater than or equal to` the specified value. | |gt |Float |Optional |100.0 |The filter range for the "range" operator, filtering data where the "field" value is `greater than` the specified value. | |lte |Float |Optional |500.0 |The filter range for the "range" operator, filtering data where the "field" value is `less than or equal to` the specified value. | |lt |Float |Optional |500.0 |The filtering range of the "range" operator. It filters data where the "field" value is `less than` a specified value. | |conds |array |Condition required |See the example below. |All operators except "range" require the conds condition.|\ | | | | ||\ | | | | |* The "must" and "must_not" operators require a set of filter values.|\ | | | | |* The "and" and "or" operators require nested filtering conditions. | **Filter Operator (op) Examples** |**Operator** |**Filtered field (field) type** |**Operator Description** |**Example** | |---|---|---|---| |must |* int32|Conditional Expression Parameter Syntax:|**Example 1:** the region must be "cn" or "sg".|\ | |* int64||```JSON|\ | |* string|* `op`: pass in "must"|{|\ | |* Bool|* `field`: pass in the target field for filtering| "op": "must",|\ | |* array(int32)|* `conds`: pass in a list of filter values, where the elements must have the same type as the field| "field": "region",|\ | |* array(int64)|| "conds": ["cn", "sg"]|\ | |* array(string) |When `field` is a single integer or string, it means **"is in"** , i.e., the field's value is in the provided filter list.|}|\ | | |When `field` is a list of integers or strings, it means **"intersects with"** , i.e., the target field and the filter list have common elements. |```|\ | | | ||\ | | | |**Example 2:** the status value must be one of 1, 2, or 3.|\ | | | |```JSON|\ | | | |{|\ | | | | "op": "must",|\ | | | | "field": "status",|\ | | | | "conds": [1, 2, 3]|\ | | | |}|\ | | | |```|\ | | | ||\ | | | | | |must_not |* int32|Conditional expression parameter format:|The status value cannot be 1, 2, or 3.|\ | |* int64||```JSON|\ | |* string|* `op`: Enter "must_not"|{|\ | |* Bool|* `field`: Enter the target field to filter| "op": "must_not",|\ | |* array(int32)|* `conds`: Enter a list of filter values. The element type must match the field type.| "field": "status",|\ | |* array(int64)|| "conds": [1, 2, 3]|\ | |* array(string) |When `field` is a single integer or string element, it means **"does not belong to"** , i.e., the target field value is not in the filter list.|}|\ | | |When `field` is a list of integers or strings, it means **"no intersection"** , i.e., the elements in the target field and the filter list do not overlap. |```|\ | | | ||\ | | | | | |range |* int32|Conditional expression parameter syntax:|**price \>= 100.0**|\ | |* int64||```JSON|\ | |* Float |* `op`: enter "range"|{|\ | | |* `field`: enter the target field to filter| "op": "range",|\ | | |* `gt`: Greater than| "field": "price",|\ | | |* `gte`: Greater than or equal to (for integers only)| "gte": 100.0|\ | | |* `lt`: Less than|}|\ | | |* `lte`: Less than or equal to (for integers only)|```|\ | | |||\ | | |The value of the target numeric field must be within the specified range.|**price is between 100.0 and 500.0**|\ | | |Use `gte` (greater than or equal to), `gt` (greater than), `lte` (less than or equal to), and `lt` (less than) to define a numeric range. |```JSON|\ | | | |{|\ | | | | "op": "range",|\ | | | | "field": "price",|\ | | | | "gte": 100.0,|\ | | | | "lt": 500.0|\ | | | |}|\ | | | |```|\ | | | | | |time_range|The filtered field must have the 'New Arrival Time' property. Supported:|Conditional expression parameter syntax:|**Published after 2026**|\ | |||```JSON|\ | |* Int64: UNIX timestamp|* `op`: pass "time_range"|{|\ | |* String: ISO 8601 time format; or "YYYY\-Mm\-DD hh:mm:ss.sss" format|* `field`: pass the target field for filtering| "op": "time_range",|\ | ||* `gt`: later than| "field": "publish_time",|\ | |For details, see [Review Reserved Field Attributes](/docs/85296/1873471#bb64f3ff) |* `lt`: earlier than| "gt": 1767196800000,|\ | | ||}|\ | | |The target filter field must be within a time range. Supported input values for `gt` and `lt`:|```|\ | | |||\ | | |* Absolute time condition (fixed period): Pass in a UNIX timestamp (int) **accurate to the millisecond**.|**Published within the last month**|\ | | |* Relative time condition (relative to the request time): Pass in the expression "now[+/\-][value][d/h/m/s]" as a string. Relative time expression examples:|```JSON|\ | | | * `now+2d`: A point in time 2 days after the current time.|{|\ | | | * `now-24h`: A point in time 24 hours before the current time.| "op": "time_range",|\ | | || "field": "publish_time",|\ | | | | "gt": "now-30d",|\ | | | |}|\ | | | |```|\ | | | | | |geo_distance |In your item dataset, this field must be an **object (Object)** and must contain two subfields named `latitude` (latitude) and `longitude` (longitude). A list of objects (List of Object) is not supported. |**Parameter format for condition expressions:** |**Example 1: Use the user’s current location as the reference point**|\ | | ||This example filters all items whose `store_location` field is within 5 km of the user’s current location (passed via `context.location`).|\ | | |* `op`: Pass in "geo_distance".|```JSON|\ | | |* `field`: The name of the field to apply geolocation filtering to. The field type must be an object (Object) that contains the `latitude` and `longitude` subfields.|{|\ | | |* `center`: The reference point for distance calculation, in the form of a `"{longitude},{latitude}"` string. If this parameter is not provided, the system uses `context.location` in the Search API request as the reference point by default.| "query":{|\ | | |* `radius`: The distance range, in the form of a `"{value}{unit}"` string. For example, `"5km"` or `"10000m"`. | "text":"local restaurant"|\ | | | | },|\ | | | | "context":{|\ | | | | "location":{|\ | | | | "longitude":"116.412138",|\ | | | | "latitude":"39.914912"|\ | | | | }|\ | | | | },|\ | | | | "filter":{|\ | | | | "op":"geo_distance",|\ | | | | "field":"store_location",|\ | | | | "radius":"5km"|\ | | | | }|\ | | | |}|\ | | | |```|\ | | | ||\ | | | | |\ | | | |**Example 2: Use the specified latitude and longitude as the reference point**|\ | | | |This example filters all items whose `store_location` field is within 10000 m of the specified latitude and longitude coordinates (`121.4737,31.2304`).|\ | | | |```JSON|\ | | | |{|\ | | | | "query":{|\ | | | | "text":"hotels near this location"|\ | | | | },|\ | | | | "filter":{|\ | | | | "op":"geo_distance",|\ | | | | "field":"store_location",|\ | | | | "center":"121.4737,31.2304",|\ | | | | "radius":"10000m"|\ | | | | }|\ | | | |}|\ | | | |```|\ | | | | | |and |Not applicable |Logical operator: Finds the intersection of multiple conditions for logical queries. | |\ | | | |```JSON|\ | | | |{|\ | | | | "op": "and",|\ | | | | "conds": [ // Condition list, supports nested and/or/must/must_not/range operators.|\ | | | | {|\ | | | | "op": "must",|\ | | | | "field": "status",|\ | | | | "conds": [1, 2, 3]|\ | | | | },|\ | | | | {|\ | | | | ... // Supports combining any number of conditions (>=1).|\ | | | | }|\ | | | | ]|\ | | | |}|\ | | | |```|\ | | | ||\ | | | | | |or |Not applicable |A logical operator that finds the union of multiple conditions for logical queries. | |\ | | | |```JSON|\ | | | |{|\ | | | | "op": "or",|\ | | | | "conds": [ // Condition list, supports nested and/or/must/must_not/range operators.|\ | | | | {|\ | | | | "op": "must",|\ | | | | "field": "status",|\ | | | | "conds": [1, 2, 3]|\ | | | | },|\ | | | | {|\ | | | | ... // Supports combining any number of conditions (>=1).|\ | | | | }|\ | | | | ]|\ | | | |}|\ | | | |```|\ | | | ||\ | | | | | #### Context |**Parameter** |**Example value** |**Description** | |---|---|---| |`location` |```JSON|1. Accepts user location data for configuring item boosting and de\-boosting based on geography. The input data must be a string.|\ | |{|2. If the input is invalid, the platform ignores the LBS configuration. |\ | | "latitude": "12.12",| |\ | | "longitude": "-13.43"| |\ | |}| |\ | |```| |\ | | | | --- ## Response Parameters #### Request and Response |**Parameter** |**Type** |**Example value** |**Description** | |---|---|---|---| |`request_id` |string |25ee998a\-5462\-9385\-9f18\-035f1da7a6e5 |A unique identifier for a request. Use this ID to quickly track its execution. | |`spell_correction` |SpellCorrection | |Returns the correction mode configuration and the corrected search term suggestion when search term correction is triggered. | |`result` |Result | |Search result. | |`total_items` |int |200 |The total number of search results returned | #### Data Structure ##### SpellCorrection |**Parameter** |**Type** |**Example value** |**Description** | |---|---|---|---| |`mode` |string |"suggestion_only" |The search correction mode configured in the console, which returns two enumerated values:|\ | | | ||\ | | | |* "suggestion_only": Only returns suggested search terms, and uses the original user\-uploaded term for the search.|\ | | | |* "auto": Automatically uses the corrected search term for the search.|\ | | | |* "off": The search term correction feature is not enabled in the console. | |`corrected_query` |string |"4K display TV" |The corrected search query. An empty value indicates that no search terms needed correction. | ##### Result |**Parameter** |**Type** |**Example value** |**Description** | |---|---|---|---| |`search_results` |array<[SearchResult](/docs/85296/1544974#7a70ec00)\> | |List of search results. | |`total_items` |int |300 |The total number of search results, which is the number of candidates. You can adjust the maximum number of items to recall in the application configuration, but the final count depends on the total amount of relevant data in the data source. | ##### SearchResult |**Parameter** |**Type** |**Example Value** |**Description** | |---|---|---|---| |`_id` |string |"WSHOE001"|The primary key for this data in data storage is the value from the field with the **unique identifier** attribute in your uploaded content. |\ | | |"video_900026" | | |`display_fields` |object |Image and text data:|Other fields in the dataset are the same as the fields from the data import. |\ | | |```Plain Text| |\ | | |{ | |\ | | | "item_id": "WSHOE001", | |\ | | | "title": "舒适通勤尖头浅口高跟鞋 - 黑色", | |\ | | | "category": "女士高跟鞋", | |\ | | | "status": 1| |\ | | |}| |\ | | |```| |\ | | || |\ | | |Video data:| |\ | | |```Plain Text| |\ | | |{| |\ | | | "_video_generated_summary":"叶文洁为防止发射信号的事被雷志成上报，制造故障引其和杨卫宁下悬崖并锯断绳子将二人杀害，史强因核辐射住院，叶文洁回忆起在齐家屯感受到人间温暖及后续经历，同时作战中心截获三体信息，汪淼和史强进入三体游戏遇到 1379 号监听员。"| |\ | | | "video generated title":"叶文洁承认杀害雷志成",| |\ | | | "content_id":"video_900026",| |\ | | | "content_type":"video",| |\ | | | "parent_content_id":"collection_342534",| |\ | | | "sequence_index":26,| |\ | | | "video_url":["http://xxxx.mp4"],| |\ | | | "my_title":"三体第二十六集"| |\ | | |}| |\ | | |```| |\ | | | | | |`score` |Float |0.852 |The merged score from multiple recall channels | |`recall_info`|array|```JSON|Records the information for each recall path.|\ | | |[|recall_reason indicates the recall method. It currently supports "text" for text recall and "vlm" for image recall.|\ | | | {|recall_score indicates the relevance score for the recall path. |\ | | | "recall_reason": "text'",| |\ | | | "recall_score": 0.7| |\ | | | },| |\ | | | {| |\ | | | "recall_reason": "vlm'",| |\ | | | "recall_score": 0.5| |\ | | | }| |\ | | |]| |\ | | |```| |\ | | | | | |`rerank_info`|object |```JSON|Reranking information for recalled data, which indicates whether the data has been reranked and its rerank score. The `rerank_score` field does not exist for data that has not been reranked. |\ | | |{| |\ | | | "is_reranked": true,| |\ | | | "rerank_score":0.993242| |\ | | |}| |\ | | |```| |\ | | || |\ | | | | | ## API Example ### Using the Associated Image\-Text Dataset #### Request Example ```JSON curl -X POST 'https://aisearch.cn-beijing.volces.com/api/v1/application/${application_id}/search/${secene_id}' \ -H 'Content-Type: application/json' \ -H 'Authorization: Bearer ' \ -d '{ "query": { "text": "女鞋春夏" }, "page_number": 1, "page_size": 5, "user": { "_user_id": "ldy_199432" }, "filter": { "op": "must", "field": "status", "conds": [1, 2] }, "dataset_id": "106265704", "context": { "location":{ "latitude": "12.12", "longitude": "-13.43" } } }' ``` #### Response Example ```JSON { "request_id": "25ee998a-5462-9385-9f18-035f1da7a6e5", "result": { "search_results": [ { "_id": "WSHOE001", "display_fields": { "item_id": "WSHOE001", "title": "舒适通勤尖头浅口高跟鞋 - 黑色", "category": "女士高跟鞋", "status": 1, "images": [ { "image_url": "https://example.com/images/womens_pump_black.jpg" } ] } }, { "_id": "WSHOE002", "display_fields": { "item_id": "WSHOE002", "title": "轻便网面跑步运动鞋 - 粉色拼灰色", "category": "女士运动鞋", "status": 1, "images": [ { "image_url": "https://example.com/images/womens_sneaker_pink.jpg" } ] } }, { "_id": "WSHOE003", "display_fields": { "item_id": "WSHOE003", "title": "波西米亚风格坡跟凉鞋 - 米色", "category": "女士凉鞋", "status": 2, "images": [ { "image_url": "https://example.com/images/womens_sandal_beige.jpg" } ] } } ], "spell_correction": { "mode":"off" } "total_items": 300 } } ``` ### Using the Associated Video Dataset #### Request Example ```JSON curl -X POST 'https://aisearch.cn-beijing.volces.com/api/v1/application/${application_id}/search/${scene_id}' \ -H 'Content-Type: application/json' \ -H 'Authorization: <火山鉴权的一系列相关header>' \ -d '{ "query": { "text": "科幻片" }, "page_number": 1, "page_size": 5, "user": { "_user_id": "ldy_199432" }, "filter": { "op": "must", "field": "director", "conds": ["郭帆"] }, "dataset_id": "106265704", "context": { "location":{ "latitude": "12.12", "longitude": "-13.43" } } }' ``` #### Response Example ```JSON { "request_id": "25ee998a-5462-9385-9f18-035f1da7a6e5", "result": { "search_results": [ { "_id": "video_900026", "display_fields": { "content_id": "video_900026", "content_type": "video", "parent_content_id": "collection_342534", "sequence_index": 26, "video_url": [ "http://xxxx.mp4" ], "my_title": "三体第二十六集" } }, { "_id": "video_900027", "display_fields": { "content_id": "video_900027", "content_type": "video", "parent_content_id": "collection_342534", "sequence_index": 27, "video_url": [ "http://xxxx.mp4" ], "my_title": "三体第二十七集" } }, { "_id": "video_900028", "display_fields": { "content_id": "video_900028", "content_type": "video", "parent_content_id": "collection_342534", "sequence_index": 28, "video_url": [ "http://xxxx.mp4" ], "my_title": "三体第二十八集" } } ], "spell_correction": { "mode":"auto", "corrected_query":"三体电视剧集" } "total_items": 300 } } ``` ## Error Handling ### Error Code |**HTTP Status Code** |**Error Code** |**Error Message** |**Meaning** | |---|---|---|---| |400 |`MissingParameter` |The request is missing the following parameter: {parameters} |A required parameter is missing in the request. | |400 |`InvalidParameter` |A parameter in the request is not valid: {parameters} |The parameter is invalid, for reasons such as an incorrect data type, invalid length, an invalid enum value, or failure to match a regular expression. | |400 |`InvalidRequestFormat` |The request cannot be parsed in JSON format. |Failed to parse the request body. This usually occurs when the request body has an invalid format, such as an illegal JSON format or a structure that does not match the request definition. Refer to the corresponding API documentation to check the request structure. | |400 |`InvalidAction` |The specified action is invalid: {Action} |The Action in the request path does not exist or is temporarily unavailable:|\ | | | ||\ | | | |* `search`: Unable to perform a search. This may occur if the index has not taken effect after the dataset was associated, or if the application is not associated with a dataset for the corresponding dataset_id. | For other common error codes, see the [Common Error Codes](/docs/85296/1544941)documentation. :::tip tip To explore more Viking AI search capabilities, visit the [Viking AI Search Engine, an all-in-one intelligent engine for enterprise search, recommendation, and Q&A](https://www.volcengine.com/product/AI-Search-Rec). :::