> ## Documentation Index
> Fetch the complete documentation index at: https://aisa.one/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Search Videos Within a WeChat Channel

> Search Videos Within a WeChat Channel

Search Videos Within a WeChat Channel

**Pricing** — billed per successful call; the final charge scales by your plan multiplier. Charged only when the upstream response body `code` is `200`.

<Note>
  Response structure is not yet fully documented — **refer to the actual API response**.
</Note>

## Example

```bash theme={null}
curl -X POST "https://api.aisa.one/apis/v1/tikhub/wechat_channels/v2/fetch_search_channel_videos" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{}'  # see request schema below
```


## OpenAPI

````yaml openapi/tikhub.json POST /tikhub/wechat_channels/v2/fetch_search_channel_videos
openapi: 3.0.0
info:
  title: TikHub API
  description: >-
    Multi-platform social media data — Douyin, TikTok, Bilibili, Kuaishou,
    Weibo, Xiaohongshu, Zhihu, and commercial (Xingtu/Huahuo/Billboard)
    endpoints.
  version: 1.0.0
servers:
  - url: https://api.aisa.one/apis/v1
security:
  - bearerAuth: []
paths:
  /tikhub/wechat_channels/v2/fetch_search_channel_videos:
    post:
      tags:
        - TikHub - Wechat_channels
      summary: Search Videos Within a WeChat Channel
      description: >-
        WeChat Channels (v2). Pricing and billing notes describe the upstream
        provider. Use the AIsa price quote for gateway calls.


        ### Purpose

        - In-channel search: search a creator's published videos by keyword
        **within that channel only**.

        - **Single page per call.** When `continue_flag=1` and `next_cursor` is
        non-empty, pass `next_cursor` back verbatim as the `cursor` parameter to
        get the next page.

        - `next_cursor` is an **opaque string** — do not parse or concatenate
        it. Upstream seals it (older versions leaked search-session state in
        plaintext).

        - ⏱️ Due to WeChat server latency, this endpoint responds slowly; please
        set your client request timeout to 30 seconds — a timeout that is too
        small may result in being billed without receiving the response.

        - ⚠️ Large-integer ID precision: IDs such as `doc_id` in the response
        are 64-bit big integers beyond JavaScript's safe-integer range (2^53-1).
        Always receive / pass such IDs as **strings** (parse JSON with
        json-bigint or read them as text), never through JS `Number`. Swagger UI
        rounds the trailing digits of huge integers in its docs view — this is
        expected and does not affect the actual data returned by the API.


        ### Parameters

        - username: Creator finder username (`v2_…@finder` format). E.g.
        `v2_060000231003b20faec8c6e4811dc1d4c602ee30b0771bbcf220c67926bb76ab7702ac335a53@finder`
        (People's Daily).

        - keyword: Keyword to search within the channel (1-100 chars after
        trimming). For example, `news`.

        - raw: Optional, default True. True=raw search response;
        False=simplified parsed structure.


        ### Return

        - In-channel search hit list


        ### Response structure & JSON Path

        #### `raw=false` (simplified, snake_case):

        - `$.data.username` / `$.data.keyword` — input echo

        - `$.data.count` — hits on this page (e.g. 20)

        - `$.data.items[]` — hit array, per item (N is the index):
            - `$.data.items[N].type` — result type (e.g. `video`)
            - `$.data.items[N].doc_id` — document id
            - `$.data.items[N].title` — title (search highlight markers stripped)
            - `$.data.items[N].desc` / `.thumb_url` / `.username` / `.nickname` — description / thumbnail / author (**may be null for some items**)
        - `$.data.continue_flag` — has more (1=yes)

        - `$.data.next_cursor` — pagination cursor (opaque string; pass it back
        as `cursor` for the next page, `null` when there is no next page)

        #### `raw=true` (full raw response, camelCase):

        - `$.data.query` → `keyword`; `$.data.offset` — next page offset (e.g.
        20)

        - `$.data.data[]` — result group array; hit items are nested at
        `$.data.data[N].subBoxes[M].items[]` (or possibly
        `$.data.data[N].items[]`), flattened into `items[]` by the simplified
        layer

        - `$.data.continueFlag` → `continue_flag`

        - `$.data.next_cursor` — same as `raw=false`: opaque pagination cursor
        (injected at top level). Older versions exposed `searchID` and other
        search context in plaintext; upstream no longer returns it


        ### Tip

        - To get media (download address / decode_key) of a hit video: this
        endpoint does not return `media` directly; feed the hit into
        `fetch_video_detail` to get `media` (`url` / `url_token` /
        `decode_key`). See `fetch_video_detail` docs for the chain and
        decryption notes.


        API references: `fetch_video_detail` = `POST
        /tikhub/wechat_channels/v2/fetch_video_detail` (operationId:
        `post_tikhub_wechat_channels_v2_fetch_video_detail`).
      operationId: post_tikhub_wechat_channels_v2_fetch_search_channel_videos
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/FetchSearchChannelVideosRequest'
        required: true
      responses:
        '200':
          description: >-
            Successful response. The outer object follows `ResponseModel`; the
            endpoint-specific payload is in `data`.


            ### Return

            - In-channel search hit list


            ### Response structure & JSON Path

            #### `raw=false` (simplified, snake_case):

            - `$.data.username` / `$.data.keyword` — input echo

            - `$.data.count` — hits on this page (e.g. 20)

            - `$.data.items[]` — hit array, per item (N is the index):
                - `$.data.items[N].type` — result type (e.g. `video`)
                - `$.data.items[N].doc_id` — document id
                - `$.data.items[N].title` — title (search highlight markers stripped)
                - `$.data.items[N].desc` / `.thumb_url` / `.username` / `.nickname` — description / thumbnail / author (**may be null for some items**)
            - `$.data.continue_flag` — has more (1=yes)

            - `$.data.next_cursor` — pagination cursor (opaque string; pass it
            back as `cursor` for the next page, `null` when there is no next
            page)

            #### `raw=true` (full raw response, camelCase):

            - `$.data.query` → `keyword`; `$.data.offset` — next page offset
            (e.g. 20)

            - `$.data.data[]` — result group array; hit items are nested at
            `$.data.data[N].subBoxes[M].items[]` (or possibly
            `$.data.data[N].items[]`), flattened into `items[]` by the
            simplified layer

            - `$.data.continueFlag` → `continue_flag`

            - `$.data.next_cursor` — same as `raw=false`: opaque pagination
            cursor (injected at top level). Older versions exposed `searchID`
            and other search context in plaintext; upstream no longer returns it
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ResponseModel'
components:
  schemas:
    FetchSearchChannelVideosRequest:
      properties:
        username:
          type: string
          maxLength: 256
          minLength: 10
          pattern: ^v2_[0-9a-fA-F]+@finder$
          title: Username
          description: Creator finder username (v2_…@finder format
          example: >-
            v2_060000231003b20faec8c6e4811dc1d4c602ee30b0771bbcf220c67926bb76ab7702ac335a53@finder
        keyword:
          type: string
          maxLength: 100
          minLength: 1
          title: Keyword
          description: Keyword to search within this channel (1-100 chars
          example: news
        cursor:
          title: Cursor
          description: >-
            Pagination cursor. Leave empty for the first page, then pass the
            previous response next_cursor verbatim. The cursor is opaque; do not
            parse or concatenate it.
          nullable: true
          type: string
          maxLength: 4096
        raw:
          type: boolean
          title: Raw
          description: True=raw response; False=simplified parsed structure
          default: true
      type: object
      required:
        - username
        - keyword
      title: FetchSearchChannelVideosRequest
    ResponseModel:
      properties:
        code:
          type: integer
          title: Code
          description: HTTP status code
          default: 200
        request_id:
          title: Request Id
          description: Unique request identifier
          nullable: true
          type: string
        message:
          type: string
          title: Message
          description: Response message (EN-US)
          default: Request successful. This request will incur a charge.
        message_zh:
          type: string
          title: Message Zh
          description: Response message (ZH-CN)
          default: 请求成功，本次请求将被计费。
        support:
          type: string
          title: Support
          description: Support message
          default: 'Discord: https://discord.gg/aMEAS8Xsvz'
        time:
          type: string
          title: Time
          description: The time the response was generated
        time_stamp:
          type: integer
          title: Time Stamp
          description: The timestamp the response was generated
        time_zone:
          type: string
          title: Time Zone
          description: The timezone of the response time
          default: America/Los_Angeles
        docs:
          title: Docs
          description: Link to the API Swagger documentation for this endpoint
          nullable: true
          type: string
        cache_message:
          title: Cache Message
          description: Cache message (EN-US)
          default: >-
            This response is cached and accessible via the URL below for 24
            hours at no extra cost. The cache is for request tracing only — it
            doesn't affect the API's data freshness and won't be returned
            through the API again.
          nullable: true
          type: string
        cache_message_zh:
          title: Cache Message Zh
          description: Cache message (ZH-CN)
          default: >-
            本次响应已缓存，可通过下方 URL 直接查看，有效期 24
            小时，访问缓存链接无额外费用。缓存仅用于请求溯源，不影响接口数据的时效性，也不会再次通过接口返回。
          nullable: true
          type: string
        cache_url:
          title: Cache Url
          description: The URL to access the cached result
          nullable: true
          type: string
        router:
          type: string
          title: Router
          description: The endpoint that generated this response
          default: ''
        params:
          title: Params
          description: The parameters used in the request
          default: {}
        data:
          title: Data
          description: >-
            Endpoint-specific response payload. The official shared model does
            not declare a fixed type or field set. Use the operation's
            documented return fields or response example; an operation-specific
            schema, when present, describes the documented fields. Undocumented
            fields must be read from the actual response rather than assumed.
          nullable: true
      type: object
      title: ResponseModel
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: AIsa API key. Get yours at https://aisa.one

````

This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.