X video transcription
X Video to Text
Speech inside videos posted on X is difficult to search and reuse. Submit a publicly accessible video post to Videosays for clean transcript text and timestamped subtitles.
Public x.com and twitter.com video posts · Plain text · Timeline · TXT / SRT / VTT
@globalbrief
Full launch clip: the lead addresses release timing, availability, and data handling.
Post copy stays as context; spoken video is transcribed separately
Spoken in the video
The first phase will open to existing users without an extra application.
Data will not be used for training unless a user actively opts in.
Countries and timing will be published in documentation next week.
Post copy is the publisher’s framing; spoken video is the actual statement
Videos on X often enter public conversation through a headline, quote post, or reply. Keep the post author, the person speaking in the video, and the timestamp of each statement separate so a paraphrase does not become a quote.
SOURCE CHAIN
Keep post framing separate from the words spoken in the video
Original post
Keep the public X post URL and publisher information
Post copy
Treat it as framing and context rather than video captions
Spoken statement
Extract what the speaker actually says in the native video
Time position
Keep timestamps so each statement can be checked in the source clip
Spoken-video timeline
Review what was actually said at each point in the video
The first phase opens to existing users without an extra application
SpokenData is not used for training unless the user actively opts in
ConditionCountries and timing will be published in documentation next week
ScheduleFuture changes should be checked against the published documentation
VerifyX videos can become these research records
News and live footage
Event, speaker, exact words, timestamps, and updates
Interviews and opinion
Questions, answers, evidence, rebuttals, and quotes
Product launches and demos
Features, scope, schedule, limits, and commitments
Sports and live replays
Key moments, commentary, reactions, and disputes
Build a claim—quote—source verification table
Available to everyone now
First phase opens to existing users
00:14
Overstated
Data is used for training
Not unless the user opts in
00:47
Conditional
Global launch next week
Countries and timing will be announced next week
01:32
Unconfirmed
VERIFY BEFORE REUSE
A transcript helps verification; it does not replace it
Videosays extracts speech and preserves timestamps. For news, disputes, health, finance, or expert claims, also inspect the full source video, official documentation, and other reliable sources.
Videosays Skill
Extract X video subtitles with the Videosays Skill and an AI agent
X video-to-text FAQ
Which X links are supported?
Videosays supports publicly accessible posts containing native video on x.com and twitter.com. Photo-only, private, deleted, or region-restricted posts cannot be processed.
Is the X post text used as the transcript?
No. Videosays transcribes the media audio. Post text is retained only as title and context.
Can I export X video subtitles?
Yes. When timeline segments are available, you can export TXT, SRT, or VTT.
Can an AI agent automatically extract X or Twitter video subtitles?
Yes. The Videosays Skill lets an agent submit a public X video post, check status, and retrieve transcript text, SRT, or VTT for an event brief.