WeChat API

On June 13, 2026 the legacy /api/v1/wechat_channels/* and /api/v1/wechat_mp/web/* series were retired. Move 公众号 calls to /api/v1/wechat_mp/v2/* (e.g. fetch_mp_article_detail_json/html → fetch_article_detail; fetch_mp_article_read_count → fetch_article_stats; fetch_search_* → the unified 搜一搜 endpoint), and 视频号 calls to /api/v1/wechat_channels/v2/*. The V2 routes use a clean version-numbered naming convention.

fetch_video_detail returns the video's media address plus a decode_key. Provide at least one input, in priority order object_id > export_id > share_url: object_id (numeric) comes from fetch_user_videos / fetch_collection_videos; export_id comes from search results (starts with export/, expires quickly — pair it with object_nonce_id / feedNonceId to improve hit rate); share_url is a weixin.qq.com/sph/… link used only when the others are empty. Set raw=false for a simplified, download-ready structure. The bare url may not open (anti-hotlinking) — use full_url (= url + url_token). If the MP4 won't play it's encrypted: decrypt with the decode_key from the SAME response — every request returns a new link + key pair.

1// fetch_video_detail (V2), raw=false → $.data.media is one object
2const m       = data.media;
3const fullUrl = m.full_url;     // = m.url + m.url_token (anti-hotlinking)
4const key     = m.decode_key;   // decrypt the MP4 with this
5// each request returns a NEW link + decode_key — use a matching pair from ONE response
6// raw=true instead → $.data.objects[0].objectDesc.media[0].{url, urlToken, decodeKey}

WeChat Channels video decryption tool

WeChat servers are slow, so set your client request timeout to about 30s — too small a timeout can bill you for the request without returning a response. Play and download counts (播放量/下载量) are not available because WeChat does not expose them publicly.

The 搜一搜 endpoint (wechat_search/v2/fetch_search) searches BOTH 公众号 and 视频号 in a single call — the recommended starting point for discovery. WeChat search does not paginate: to get more results, send the same request again (some results may repeat). Search runs on limited backend resources, so a first attempt can return partial results — retry a few times if needed.

On WeChat Official Accounts, an article's read count (阅读数) is the key metric — it reflects how many people actually opened the piece inside a closed messaging app, a stronger attention signal than a public like count. fetch_article_stats (文章互动数据) returns the interaction data, and fetch_article_detail returns full article content for parsing, AI, and archiving.

No. Create a free TikHub account, generate an API key, and call REST endpoints directly — no WeChat developer application, app review, or OAuth.

搜一搜 universal search across 公众号 and 视频号; Official Account (公众号) article content, interaction stats, comments, related articles, and ads; and Channels (视频号) profiles, videos, collections, live history, comments, and video parse/download.

Pay-as-you-go from $0.001 per request, with automatic daily volume discounts down to $0.0005 at 30,000+ requests/day. New accounts get $0.05 free credit (~50 requests) — no credit card required.

TikHub only returns publicly available WeChat data — content any visitor can see without an account. It does not bypass authentication or access private information. WeChat is a Chinese platform, so the relevant constraints are its own Terms of Service and China's data and cross-border rules rather than US case law; automated access can lead to account or IP blocks, which TikHub's infrastructure absorbs for you. You remain responsible for how you store, process, and use the data — especially any personal data and any onward transfer across borders.