Benchmarks · YouTube
The YouTube transcript scrapers people use most, measured on the same requests
We ran the same requests through YouTube Transcript Scraper and through the most used YouTube transcript scrapers on the Apify Store, then checked what came back against the source. Measured on October 10, 2026. Alternatives = the most used scrapers for the same platform on the Apify Store, ranked by monthly users (Apify public statistics). Scrapers change: these are the results on that date, and we keep the raw runs.
Open YouTube Transcript Scraper on ApifyTranscripts cost $5.00 per 1,000 on every Apify plan, plus $0.005 per run at the default memory; videos without captions are free.
What came back
| Test | Quarra | Alternatives |
|---|---|---|
| 10 latest videos of @TED and of @MrBeast, language left emptyThe most used YouTube transcript scraper. Counted on the 12 long videos both returned; its 10 also took in 8 Shorts. The Arabic tracks are translations attached to these English videos. | 12 of 12 in the spoken language | Arabic for 10 of 12 |
| The 5 long @MrBeast videos, English askedThese captions are styled with an outline, which YouTube stores as a second copy of each line. Their text also kept invisible characters. | Each caption line once | Every line twice |
| Run time, 10 latest videos of a channel@TED, then @MrBeast. Run times as recorded by Apify: ours are the median of about 20 runs each, with comment counts and subscribers; theirs, one run each. Their 10 included Shorts, ours are the Videos tab. | 16.8 s and 24.7 s | 18.1 s and 23.7 s |
| Videos per channel on Apify's free planAs its run log says on a free account. | Up to 10,000 | 10 |
Where we are not ahead
- Run time: the median of about 20 runs of ours took 16.8 s for @TED and 24.7 s for @MrBeast, one run of the most used alternative 18.1 s and 23.7 s. Close, not clearly faster.
YouTube transcript API in 2026: what the official API allows, and what to use insteadHow to download the transcripts of every video on a YouTube channelAll benchmarks and how we measure