Pricing

from $5.00 / 1,000 caption transcript (字幕逐字稿)s

Try for free

Go to Apify Store

中文影片逐字稿 · Chinese Video Transcript (AI fallback)

Try for free

Get transcripts from YouTube videos - Chinese first. Only ~25-37% of Chinese videos have captions; this Actor transcribes the rest with AI.

Pricing

from $5.00 / 1,000 caption transcript (字幕逐字稿)s

Rating

0.0

(0)

Developer

Blue59

Actor stats

Bookmarked

Total users

Monthly active users

2 days ago

Last modified

中文影片逐字稿 — 沒字幕也能抓逐字稿

Get the transcript even when the video has no captions. 以中文為核心,也支援其他語言 / Chinese-first, works for any language.

Other transcript scrapers fail on most Chinese videos, because most Chinese videos simply have no captions to scrape. We measured it: only ~25–37% of Chinese YouTube videos have any caption track, and auto-captions are almost never generated for Chinese. This Actor scrapes the captions when they exist — and transcribes the audio with AI when they don't.

你會拿到什麼 / What you get

📝 完整逐字稿 — 有字幕就抓字幕（1–3 秒／支），沒有就 AI 轉錄。Full transcript either way.
🇹🇼 中文優先 — 繁中 → 簡中 → 通用中文 → 英文，依序挑選最合適的字幕軌。Traditional Chinese first, configurable.
✅ 品質過濾 — 每筆結果都經過內容檢查：太短、錯誤頁、抓錯語言軌，一律標為失敗而非混入有效資料。Quality-gated: bad results are reported as failures, never silently returned.

範例輸出 / Example output

{
  "videoId": "pDxi9MljStw",
  "url": "https://www.youtube.com/watch?v=pDxi9MljStw",
  "title": "EP103 很怕別人過得比我爽",
  "channel": "蔡康永",
  "duration": 3821,
  "uploadDate": "20260714",
  "viewCount": 284119,
  "language": "zh-Hans",
  "source": "caption:manual",
  "charCount": 26985,
  "transcript": "簡單的來講就是不放心 不放心 我不放心別人是不是在過比我爽的很多的生活…"
}

source 告訴你這筆是怎麼來的：caption:manual（人工字幕）／caption:auto（自動字幕）／ai:groq-whisper-turbo（AI 轉錄）。

輸入 / Input

欄位	說明
`videoUrls`	影片、頻道或播放清單網址（也接受 11 碼 ID）
`preferredLanguages`	語言優先序，預設繁中優先
`aiFallback`	沒字幕時用 AI 轉錄（需 Groq 金鑰）
`maxItems`	從頻道展開時的上限

AI 補位怎麼開 / Enabling AI fallback

到 console.groq.com 申請免費金鑰（免費層每日 2,000 次）
勾選 aiFallback，把金鑰貼進 groqApiKey

金鑰只在該次執行中使用，不會被儲存。Your key is used for that run only and never stored.

誰在用 / Who this is for

內容研究／競品分析 — 一次抓整個頻道的逐字稿做主題分析
AI 工作流餵料 — 給 RAG、摘要、問答系統的乾淨文字輸入
字幕與翻譯 — 拿到原文再自行翻譯或校對
無障礙 — 為沒有字幕的中文內容補上文字

技術細節 / Technical notes

字幕優先取 json3 結構化格式（最穩），退回 vtt/srt 並自動清掉時間軸標記。失敗依狀態碼分流重試（404 立即放棄、429 拉長等待、403/5xx 退避重試），錯誤分為 proxy／rate-limit／blocked／network／other 五類統計於執行報告。建議開啟代理：YouTube 會擋機房 IP。

Xigua 西瓜视频 Search Scraper - Keyword + Comments + MP4 下载

zen-studio/xigua-video-search-scraper

Search Xigua (西瓜视频) by keyword or hashtag — up to 1,000 videos with title, author, engagement stats, 1080p/720p/480p MP4 download URLs, hashtags, covers, optional comments and replies. Cross-app linkage to Douyin (抖音). 西瓜视频下载 / 爬虫.

Zen Studio

Ctrip 携程酒店评价 & Trip.com Hotel Reviews Scraper | Global

zen-studio/ctrip-hotel-reviews-scraper

Extract guest reviews from Ctrip 携程 and Trip.com: review text, overall and per-dimension ratings, stay date, room type, traveller type, reviewer country and tier, photos, video and the hotel's own reply. 61 fields per review, plus a per-hotel rating summary. Search by link, name, district or city.

Zen Studio

YouTube 抓字幕 (Transcript Scraper)

charlsondou/youtube-zhua-zi-mu-transcript-scraper

Looking for a reliable way to extract transcripts from YouTube videos? YouTube Transcript Scraper Pro has you covered. It effortlessly retrieves transcripts with modern fingerprinting to ensure high success rates. Ready to start? Let’s scrape away!

Charlson Dou

5.0

(1)

RedNote Transcript & Subtitle API – Xiaohongshu Video to Text

lance_api/xiaohongshu-video-ai-transcript-scraper-api

Extract native subtitles and convert public RedNote (Xiaohongshu/XHS) videos to text, with AI transcription used only when needed. Get timestamped segments, SRT/VTT, optional translation, video URLs, and note metadata. A matching xsec_token is required.

LanceAPI

Bilibili (哔哩哔哩) Video Scraper ✅ NO COOKIES

atomus/bilibili-scraper

Scrape Bilibili (哔哩哔哩) by keyword or URL: video search, video details, comments, UP主 profiles, danmaku (弹幕 bullet-comments), and native subtitle transcripts, plus trending. No cookies, no login. Fast structured JSON. Pay only for results returned; failed lookups are free.

Atomus APIs

5.0

(1)

Douyin 抖音 Product Search Scraper - Price, Sales & Commission

zen-studio/douyin-product-search-scraper

Search Douyin Shop (抖音商城) products by keyword and extract price, monthly sales, a 30-day sales trend, positive-review ratio, shop name and score, and affiliate commission rate. One row per product in clean JSON, ready for product research, competitor analysis, and affiliate selection.

Zen Studio

111

5.0

(1)

Ctrip 携程酒店价格 & Trip.com Hotel Price Scraper | China & Global

zen-studio/ctrip-hotel-prices-scraper

Extract every room type and rate plan from Ctrip 携程 and Trip.com: nightly price, rack rate, discount, breakfast, cancellation policy with deadlines, payment terms and rooms left. 100 fields per rate plan. Covers hotels in mainland China and worldwide. Search by hotel link, name, district or city.

Zen Studio

Hongguo Short Drama Search & Data Scraper API

spider_studio/hongguo-search

Search Hongguo short dramas by title or keyword. Extract series cards, metadata, IDs, and pagination for up to 1,000 queries through an API or Dataset.

NewLai

食べログ Tabelog 日本餐厅行情 — 按商圈看预算·评分·菜系

jpmarketdata/tabelog-restaurant-market-cn

指定一个食べログ（Tabelog，日本最大餐厅点评网站）商圈，例如银座，返回该商圈餐厅的一条汇总。包含该商圈的餐厅总数，以及从散布在整个商圈的 60 家样本得出的晚餐/午餐预算档位与中位档、评分（中位数、区间、有评分的占比）、主要菜系与最近车站。每个商圈 $0.02，需要单店记录时每家 +$0.002，零结果不收费。Area summary from 60 sampled restaurants, $0.02 per area. Unofficial.

h ichi

日本化妆品排行榜 @cosme Top 50 评分与价格

jpmarketdata/cosme-beauty-market-cn

选一个 @cosme（日本最大美妆点评网站）的分类排行，一次拿到整个 Top 50 的汇总。返回评分典型值与区间（0–7 分制）、含税标价典型值与区间、点评数中位数与最大值、品牌数、上升/下降/新上榜的数量、排行更新日与统计期间。每个分类 $0.02，无结果不收费。@cosme Japan top-50 ranking ratings and prices per category. Unofficial.

h ichi