Bilibili is one of China’s largest video platforms and a rich read on Chinese consumer and creator trends. For brand, marketing, or market research, you want video metadata — titles, creators, and engagement — as structured data rather than scrolling the app.
This guide shows how to pull Bilibili search results as JSON.
Who uses Bilibili data
Bilibili is one of China’s largest video platforms and a primary read on Chinese youth and consumer culture. Several teams use its data. Brand and marketing teams entering or watching the China market track how products and categories are discussed. Consumer-insight and trend researchers gauge which topics, formats, and creators are gaining traction. Competitive-intelligence teams watch rivals’ presence and reception. And agencies and creator-marketing teams shortlist creators by reach and engagement before a campaign.
What the data looks like
One record per video, with engagement metrics:
| Field | Description |
|---|---|
| title / bvid / url | video title, id, and link |
| author | uploader name |
| play_count | views |
| likes / favorites | engagement counts |
| danmaku_count | on-screen comment count |
| pub_date | publish date |
Note: a search that returns nothing (or is rate-limited for anonymous access) emits a single status record instead of billing for empty results — some keyword searches perform best with a login cookie.
How to get it
Run the Bilibili Video Search on Apify with a keyword. Output is JSON, CSV, or Excel, priced per video. Optional opt-in add-ons can score sentiment or extract creator social handles.
Run it on a schedule
{
"keyword": "your topic",
"max_results": 50
}Common workflows
The usual jobs are: a keyword search ranked by engagement to see which videos and creators dominate a topic; a periodic pull to track how a topic’s momentum changes over time; a creator-discovery pass that surfaces active uploaders in a niche by plays and likes; and enrichment of a China-market research set with real engagement metrics. Because the output is clean JSON or CSV, each of these drops into a spreadsheet, a dashboard, or a downstream analysis step.
Why pull it as structured data
Scrolling the app gives you an impression, not a dataset. Turning search results into structured records — title, creator, plays, likes, favorites, danmaku, and publish date on every row — lets you rank, compare, and track topics quantitatively rather than by feel, and keep a history you can chart over time.
What the engagement metrics mean
Bilibili exposes a richer engagement picture than a simple view count. Play count shows reach; likes and favorites show approval and intent to return; and danmaku — the on-screen scrolling comments the platform is known for — shows active, in-the-moment engagement that a static comment count misses. Reading these together separates videos that were merely seen from videos that genuinely landed with an audience.
Access notes and honest limits
Anonymous keyword search can be rate-limited or return nothing for some queries; when a search yields no results, the Actor emits a single status record instead of billing for empty output, which keeps runs billing-fair. For consistent keyword coverage, some queries perform best with a login cookie. These are real constraints of the source, not the tool, and the Actor is transparent about them rather than silently returning partial data.
Cost and control
Pricing is per video returned with no subscription, and optional add-ons — sentiment scoring and creator social-handle extraction — are off by default and billed only when you switch them on. You set the keyword and result cap on every run, so cost tracks exactly what you collect, and an empty search costs effectively nothing.
FAQ
What engagement metrics are included? Plays, likes, favorites, and danmaku (on-screen comments).
Can it score sentiment? Yes, via an optional add-on (billed only if enabled).
Do I pay for empty searches? No — a no-result search returns a single status record.
What engagement metrics are included? Plays, likes, favorites, and danmaku (on-screen comments), plus title, creator, and publish date.
Can it score sentiment? Yes, via an optional add-on billed only if enabled.
Do I pay for empty searches? No — a no-result search returns a single status record rather than billed items.
For teams new to the platform, the fastest way to build intuition is to run a few keyword searches in your category and sort by plays and by danmaku separately — the two rankings often disagree, and the gap tells you which topics draw views versus which spark active discussion.
Stored over time, those rankings become a momentum tracker: you can see a topic or creator rising before it peaks, which is exactly the lead time a China-market campaign needs to act rather than react.
Read together, plays, likes, favorites, and danmaku turn a keyword search into a genuine measure of what is resonating in Chinese online culture right now.
Related: explore more China and APAC data tools on The Next Gen Nexus.
We cover AI agents, automation, and the tools that make them work. Our mission is to make AI accessible to everyone.