Kuaishou & Kwai Scraper — Videos, Comments, Profiles, Search

Extract public Kuaishou (快手 / Kwai) data — built for brand/KOL-marketing agencies, e-commerce sellers tracking live-shopping creators, and China-focused social-listening teams. Granular pay-per-event pricing: you pay only for the record types you pull.

What it does

Provide any combination of three inputs:

• videoUrls — video metadata (caption, author, view/like counts, duration, tags, cover & video URLs), plus optional comments (author, text, likes, reply flag, timestamp).
• userProfiles — a Kwai creator's profile (name, bio, fan/follow/video counts, verification) plus their videos.
• searchTerms — keyword search, returning matching videos.

Every operation runs anonymously — no login or cookie required.

How it gets the data. Video detail comes from Kuaishou's server-rendered page (fast). Comments are captcha-walled for plain HTTP clients, so the actor loads the real page in a headless browser and captures the page's own signed comment requests — no captcha solving. Profiles are fetched from Kwai (Kuaishou's international brand), which server-renders profile + video data. Search uses Kwai's REST API on an anonymous visitor session. Residential proxies are required throughout.

Example input

{
    "videoUrls": [
        "https://www.kuaishou.com/short-video/3xf8enef8rg4hkc",
        "3xnzvbtkx9xefpa"
    ],
    "includeComments": true,
    "maxCommentsPerVideo": 200,
    "proxyConfiguration": {
        "useApifyProxy": true,
        "apifyProxyGroups": ["RESIDENTIAL"]
    }
}

Example output

Records are pushed to one dataset with a record_type field (video or comment). The dataset has two pre-built views — Videos and Comments — so you can export each cleanly.

A video record:

{
    "record_type": "video",
    "id": "3xq6vcny33w5ffm",
    "source_url": "https://www.kuaishou.com/short-video/3xq6vcny33w5ffm",
    "caption": "话糙理不糙！馆长怒批…",
    "author_id": "3xtszwcjndjy5ek",
    "author_name": "中国新闻社",
    "view_count": "17万",
    "like_count": 9161,
    "duration_ms": 31480,
    "video_ratio": 0.5625,
    "cover_url": "https://p2.a.yximgs.com/...",
    "video_url": "https://v4.oskwai.com/....mp4",
    "tags": ["时事"],
    "published_at": "2026-06-09T14:38:58+00:00",
    "scraped_at": "2026-06-15T09:12:00+00:00"
}

A comment record:

{
    "record_type": "comment",
    "id": "1142338051659",
    "video_id": "3xxuh8kmqmkkm2k",
    "author_id": "3xpxafb3faju7y9",
    "author_name": "程秀海567",
    "content": "好听[爱心][爱心][爱心]",
    "like_count": 1,
    "reply_count": null,
    "has_replies": false,
    "commented_at": "2026-06-16T05:46:02+00:00",
    "scraped_at": "2026-06-16T23:42:00+00:00"
}

Field notes:

• id is stable across runs — use it to deduplicate on your side.
• like_count prefers the real (un-rounded) count when Kuaishou exposes it.
• view_count is Kuaishou's localized display string (e.g. "17万" = 170k), not an integer — the web surface does not expose a raw view integer.
• Any field can be null when Kuaishou omits it for a given video.
• Timestamps are ISO-8601 UTC.

Pricing — pay-per-event

You pay only for what each run produces. Set these prices in the Apify Console under Monetization → Pay per event:

Event	Price (USD)	Charged when
actor-start	$0.001	Once per run (low by design — no surprise base fee)
video-detail	$0.006	Per video record returned
comment	$0.003	Per comment record returned
user-profile	$0.006	Per creator profile (via Kwai, no login)
user-video	$0.002	Per video in a creator's catalog (via Kwai, no login)
search-result	$0.002	Per search-result video (via Kwai, no login)

Pricing is benchmarked against the Kuaishou/Kwai actors on the Apify Store: the budget incumbents charge ~$0.002–0.003 per flat result, while the premium full-feature competitor charges ~$0.01 per video detail and ~$0.002–0.004 for bulk rows (comments, search, catalog). This actor sits in between — cheaper than the premium option, with higher reliability than the budget ones. A typical run of 1 video + 100 comments costs about $0.31.

Limitations

• Public content only. Private, age-gated, or removed videos return nothing for that ID (logged, not billed).
• Residential proxies required. Both the video page and the headless-browser comment fetch need residential IPs — datacenter IPs are blocked. Residential is the prefilled default. The browser uses a sticky session so all its connections share one IP.
• Comments need the browser path. When comments are enabled the actor launches a headless browser (heavier and slower than metadata-only runs — budget ~60–90s per video). Video-only runs (comments off) use a fast, lightweight HTTP path.
• Profiles & search use Kwai, not kuaishou.com. Kuaishou's own web profile/search routes are login-walled, so both are served via Kwai (Kuaishou's international brand, same backend), anonymously. For profiles, provide a Kwai handle or www.kwai.com/@handle URL. Note Kwai's catalog is international and overlaps with — but isn't identical to — China-domestic Kuaishou; some China-only creators may not be present.
• Comment depth. Only root (top-level) comments are returned; whether a comment has replies is flagged by has_replies. Nested replies are not expanded in v1.
• Schema drift. Kuaishou changes its internal data schema periodically (it recently moved comments to rootCommentsV2); parsers accept both old and new shapes and degrade gracefully to null rather than crash. Checked monthly.

FAQ

Where do I find a video's photo ID? It's the last segment of the URL, e.g. https://www.kuaishou.com/short-video/<photoId>. You can paste the full URL or just the ID.

Can I scrape comments without video details? Each video you list is fetched for its detail record first; set includeComments: false for cheap metadata-only runs.

Can I run this on a schedule? Yes — use Apify's Schedules feature to re-pull a creator's videos or monitor comment growth over time.

Which proxy should I use? Residential (the prefilled default). Switch groups if you start seeing empty results.

How do I export to my DB / Google Sheets? Use Apify Integrations or the Dataset API. Every field above is available via /items?format=json|csv|xlsx, and the Videos / Comments views split the two record types for you.