ae_tiktok_fetch_general_search
Searches for a general list of content matching a keyword.
Inputs:
- `keyword` (required): Search term.
- `offset` (optional, default `0`): Pagination cursor. Use the value returned by the previous response when requesting another page.
- `search_id` (optional, default empty): Pagination identifier. For a subsequent page, use the identifier from the previous response, such as `data.extra.logid` or `data.log_pb.impr_id` when present.
- `cookie` (optional): Provide a user cookie only when needed for account-based searching or when access requires it.
Returns the general search result data, including any result metadata and available video links. Access to some returned video CDN links may require the token information included in the response; follow the response's access instructions.
ae_douyin_billboard_fetch_hot_total_high_play_list
## Purpose
Retrieve the Douyin high-completion-rate leaderboard, optionally filtered by time window, keyword, and content category.
## Inputs
All inputs are optional and belong inside `request_body`:
- `page`: integer page number; default `1`.
- `page_size`: integer number of items per page; default `10`.
- `date_window`: integer time window in hours: `1`, `24`, `72`, or `168`, meaning the last hour, day, three days, or seven days. Default: `24`.
- `keyword`: string used to filter the leaderboard; default `""` means no keyword filter.
- `tags`: array of content-category filters; use `[]` for no category filter. Each object has `value` for a top-level category ID and optional `children`, an array of objects with `value` for subcategory IDs. Omitting `children` includes all subcategories under that top-level category. Obtain IDs from the content-tag lookup.
## Example
```json
{"request_body":{"page":1,"page_size":10,"date_window":24,"keyword":"","tags":[]}}
```
## Result
Returns the requested high-completion-rate leaderboard in the response's `data`, with response status and metadata. The documented schema does not specify individual leaderboard entry fields or the precise metric calculation. This is a completion-rate leaderboard, not a view-count leaderboard.
ae_douyin_search_fetch_search_suggest
## Purpose
Get suggested Douyin search phrases related to an input keyword, for search-box autocomplete or query refinement. This tool returns keyword suggestions rather than videos or search-volume statistics.
## Input
- `request_body` (required object): Contains the search-suggestion input.
- `request_body.keyword` (optional string): The term to expand. The schema default is `人工智能`. Set it explicitly to request suggestions for your intended term.
Example:
```json
{"request_body":{"keyword":"人工智能"}}
```
## Result
The response envelope includes `code` (HTTP status code), `message` and `data` (the result payload, which may be null). The documented payload includes:
- `status_code`: Result status; `0` indicates success.
- `status_msg`: Result message.
- `sug_list`: Suggested search phrases. Each item's `content` is the suggested text; `pos` describes matching character positions, and `word_record` contains suggestion metadata such as its position and text.
Additional request identifiers and metadata may be present. The detailed payload is not fixed by the response schema; inspect the actual response for available fields and distinguish the envelope's `code` from the payload's `status_code`. Suggestions are not evidence of search volume or popularity.
ae_weibo_realtime_search
## Purpose
Search Weibo for the latest posts matching a keyword, sorted by time.
## Inputs
- `query` (required string): The keyword to search for.
- `page` (optional integer, default `1`): The page of results to retrieve.
## Results
Returns search results in the response data. Results may include post content, author information, images, videos, and interaction metrics. The detailed result schema is not specified; available details may vary.
## Notes
If you play a returned video, the playback request must include `Referer: https://weibo.com/`. This is a playback requirement, not an additional search input.
## Example
`{"query":"苹果发布会","page":1}`
ae_youtube_get_video_comments
## Purpose
Retrieve a page of top-level comments for a YouTube video, with top or newest sorting and optional pagination. This tool reads comments; it does not post comments or retrieve reply threads.
## Inputs
- **video_id** (required string): The exactly 11-character YouTube video ID, for example `oaSNBz4qMQY`. Use the ID from the video's URL, not the full URL.
- **sort_by** (optional string): `top` for top comments (default), or `newest` for the latest comments.
- **language_code** (optional string): Language preference, such as `zh-CN`, `en-US`, `ja-JP`, or `ko-KR`. Default: `zh-CN`.
- **country_code** (optional string): Region code, such as `US`, `JP`, or `GB`. Default: `US`.
- **continuation_token** (optional string): Omit for the first page. To request another page for the same video, supply the comments-page token returned by the previous response.
- **need_format** (optional boolean): `true` (default) returns cleaned, simplified data; `false` returns raw data.
## Results
The response includes a `data` payload. With `need_format=true`, the documented comment payload contains a `comments` list and a page-level `continuation_token` when available. Comment entries can include:
- `comment_id`, `content`, and `published_time`: Comment ID, text, and relative publication time.
- `reply_level`: `0` for top-level comments.
- `like_count`, `reply_count`, and related display text: Likes and reply counts, which may be represented as strings.
- `author`: Channel ID, display name, channel link, avatars, and verification, creator, or artist indicators.
- `creator_thumbnail_url`: Video creator's avatar URL.
- `reply_continuation_token`: A token associated with that comment's replies; this is not the page-level token for more top-level comments and does not mean replies have been retrieved.
A single request returns one page, not necessarily every comment. Raw output has a different structure from the cleaned comment payload.
## Example
`{"video_id":"oaSNBz4qMQY","sort_by":"newest","need_format":true}`
ae_wechat_get_user_profile
## Purpose
Get a WeChat Channels creator's homepage profile and account-level interaction statistics.
## Inputs
Provide the required `request_body` object:
- `username` (required): finder username in `v2_…@finder` format, matching `^v2_[0-9a-fA-F]+@finder$`, with a total length of 10–256 characters. Use the username obtained from video data, or convert an `sph…` short ID first. A nickname or short ID cannot be supplied directly.
- `raw` (optional boolean, default `true`): `true` returns the full nested raw response; `false` returns a simplified, flattened profile using snake_case field names.
## Results
Profile information is returned in `data`. With `raw=false`, fields include:
- `username`, `nickname`, `signature`, `head_url`, and `region`: account identity, biography, avatar URL, and IP region.
- `fans_count`, `feeds_count`, `like_count`, `fav_count`, `forward_count`, and `friend_follow_count`: follower, video, like, favorite, forward, and friend-follow counts.
- `live_duration_hours`: accumulated live-streaming duration in hours.
- `original_flag`, `collection_count`, and `verify_desc`: original-content flag, number of collections, and verification description, which may be null.
- `linked_accounts`: linked official accounts with `nickname` and `type`.
With `raw=true`, profile fields are nested under `data.contact`; counts include `data.fansCount`, `data.feedsCount`, `data.feedsLikeCount`, `data.feedsFavCount`, and `data.feedsForwardCount`. IP region is at `data.ipRegionInfo.regionText`, collection count at `data.collectionListInfo.collectionCount`, and linked-account details at `data.bizInfo.info`. The raw response also includes first-page video and live-session information; it is not a complete video listing.
## Notes
Some accounts return 0 for follower or interaction counts; do not assume these values prove there is no audience or activity. This lookup is for profile data, not dedicated verification-entity/license information or collection details. Results are returned directly, without polling. Responses may be slow; a 30-second client request timeout is recommended where configurable.
ae_douyin_fetch_multi_video
Retrieve information for 1–10 Douyin video or image/text post IDs. Supply the IDs as strings in request_body. For larger lists, make separate calls with at most 10 IDs each. Deleted, private or unavailable posts may be absent; the returned count is not guaranteed to match the input count.
ae_douyin_index_fetch_topic_query
Search for topics by keyword on Douyin or Toutiao within a specified date range.
Inputs:
- keyword (required): The topic keyword to search for.
- start_date (required): Start date as a YYYYMMDD string.
- end_date (required): End date as a YYYYMMDD string.
- app_name (optional): aweme for Douyin (default), or toutiao for Toutiao.
Returns: Response status information and a data payload containing topic details and related metrics, such as topic popularity and related video counts, when available. The payload may be null, and its exact fields are not fixed by the documented response schema; use the fields actually returned.
Example: keyword="旅行", start_date="20260406", end_date="20260412", app_name="aweme".
ae_reddit_post_request
Retrieve the main content of a specified Reddit post, including its title, full body text, and attached multimedia such as images or videos. Provide the required reddit_id as a string with the t3_ prefix. This is a lookup by post ID, not a post search. The provider does not specify a response schema, so exact result field names and multimedia representations are not defined.
ae_douyin_index_fetch_relation_word
Retrieve words associated with a keyword on Douyin or Toutiao for a specified date range, including search-related words, content-related words, rankings, and relationship graph data.
Inputs:
- keyword (required): The keyword to analyze.
- start_date (required): Start date as a YYYYMMDD string; Monday is recommended.
- end_date (required): End date as a YYYYMMDD string; it must be a Sunday. A different weekday may result in empty data.
- app_name (optional): aweme for Douyin (default), or toutiao for Toutiao.
Results describe words associated with the keyword in search and content, along with rankings and graph relationships. The detailed result structure is not specified, so specific scores or graph fields are not guaranteed.
Example: keyword="美食", start_date="20260330", end_date="20260405".
ae_douyin_xingtu_get_kolid_by_uid
Look up the Xingtu visitor KOL ID (kolid) associated with a Douyin user. Supply the required uid as a string containing the Douyin user ID, for example "70452002324". You can obtain this ID from user-profile information. The response data provides the kolid lookup result; its inner structure is not specified and it may be null. This is an identifier lookup, not a full user-profile lookup.
ae_douyin_xingtu_multi_get_item_stats
Retrieve statistics and details for up to 20 Douyin videos in one request.
Inputs:
- request_body is required. Set request_body.item_ids to an array of video ID strings. Empty entries and duplicates are automatically removed; requests exceeding 20 IDs return an error.
- request_body.platform_source: optional integer identifying the platform source; defaults to 1.
- request_body.need_cover_url: optional boolean controlling whether cover URLs are returned; defaults to true.
Results:
- Matched items with video details and stats: watch_cnt (plays), like_cnt (likes), comment_cnt (comments), share_cnt (shares), favorite_cnt (favorites), interact_rate (interaction rate), and avg_watch_dura (average watch duration per viewer).
- watch_cnt includes paid promotion and must not be interpreted as organic-only plays.
- Invalid or deleted video IDs may be omitted, so fewer items may be returned than requested.
ae_youtube_get_channel_description
Retrieve a YouTube channel's description, profile details, and channel statistics.
### Inputs
- Provide `channel_id` for the initial request, or `continuation_token` from the initial response for the follow-up request. One of these is required.
- `channel_id` is typically a 24-character string starting with `UC`. If you only have a channel URL, resolve it to a channel ID first; this input is not a URL.
- `need_format` (optional, default `true`): return cleaned, simplified data. Set to `false` for raw data.
- `language_code` (optional, default `zh-CN`): preferred display language, such as `en-US` or `ja-JP`.
- `country_code` (optional, default `US`): region code, such as `US`, `JP`, or `GB`.
### Results and two-request workflow
1. Call with `channel_id` to get basic details such as channel title, handle, description, subscriber and video counts, avatar, banner, channel URLs, and a `continuation_token`.
2. Use that returned `continuation_token` in a second call to get additional About information, including creation date, social links, country, and total view count.
Complete channel information requires both requests; results are not automatically combined. With cleaned output, fields such as `subscriber_count`, `video_count`, and `view_count` can be display strings rather than numbers, and dates can be localized text. Some fields can be null or empty: for example, the first response may lack creation date and social links, while the follow-up may lack title, avatar, or banner. Retain populated details from both responses when combining them. A continuation token retrieves additional channel information; it is not a background-task identifier.
ae_weibo_hot_search_summary
Retrieve the complete Weibo hot-search list, documented as 50 entries, to see trending topics and their relative popularity.
Inputs: None. Call with an empty object ({}).
Returns: Response data containing the ranking, with each topic's rank, keyword, tag (such as Hot, Boiling, Official, or New), and heat value. Rank 0 indicates pinned content.
Notes: This is the full-list view rather than a top-10 overview. No date or filtering parameters are available. For repeated checks, caching results for 5–10 minutes is recommended.
ae_weibo_pic_search
Search Weibo for images matching a keyword, with multiple images grouped by Weibo post ID.
Inputs:
- query (required string): The search keyword.
- page (optional integer, default 1): The page of results to retrieve.
Returns image search results containing post IDs, thumbnails, original-image URLs, author information, and image counts. Thumbnail URLs are automatically converted to original-image URLs.
Example: query = "yu7", page = 1.
ae_wechat_channels_search_accounts
Find WeChat Channels creator accounts and related channel objects using a search keyword. Supply the required keyword as a string; no pagination input is needed. The response includes search data (data), a result code (code), a message (message), and recordTime. The fields within data are not specified, and message and recordTime may be null. Check the result code and message when interpreting the response.