Automatic subtitles, ready to review and publish

echoflow's automatic subtitles come from a transcript of the real spoken audio: you upload the video, the AI transcribes what is said and derives subtitles that are already synced, downloadable as SRT and VTT or burned straight into the video. In the spoken language, with nothing to configure, and editable before you publish. You get a 14-day trial on sign-up, no card.

It isn't a separate service you buy on the side: the same upload also produces the full transcript, the summary, the posts, the article and the vertical clips.

Two ways to use them

The subtitle file (.srt or .vtt) is what you upload to YouTube, hand to your editor or attach to a video player on your own site. The advantage is that it stays separate from the video: viewers can switch it on or off, and search engines can read it.

The already-captioned video is the .mp4 with the text burned into the image. It's what you need where separate subtitles don't exist — social platforms, where videos usually start muted and with no way to turn captions on. It's generated after processing, once per job, with no extra minutes. If the video was uploaded during the trial it carries the echoflow watermark; if it was uploaded on Pro it's clean.

On vertical clips, word by word

Short clips in 9:16 format come with karaoke subtitles: each word highlights as it's spoken, not in rough blocks. That's possible because speech recognition gives us the timing of every individual word rather than an estimate per line. The style is fixed in the trial and customisable with Pro.

Editing them is part of the job, not a fallback

No automatic system always gets the proper noun, the industry acronym or the technical term right. That's why the text stays editable: you fix what needs fixing and the export updates. Anyone promising perfect subtitles you never have to reread is describing a product that doesn't exist — and on a public video, one wrong name is more noticeable than ten minutes saved.

Accessibility: what we automate and what stays with you

The European Accessibility Act has made subtitles and transcripts a concrete requirement for many organisations, and the real cost isn't producing one: it's producing hundreds, across an entire archive. That part we automate.

What we cannot do is declare your content compliant. Compliance depends on the accuracy of the text, speaker identification, relevant non-speech sounds and how the video is published: it concerns the published content, not the tool that produces it. No automatic generator can guarantee it, ours included. We give you production at scale and editable text; review and verification stay with you.

Where the data is

Files and content stored in the European Union. Transcription and AI generation also use non-EU providers, listed by name on our sub-processors page, under the safeguards of Article 44 and following of the GDPR. The source file you upload is deleted within 24 hours of processing unless the workspace has set its own retention.

What we don't do

• We don't identify speakers: subtitles don't indicate who is talking in a multi-voice

conversation.

• We don't certify accuracy and don't quote percentages: results depend on the audio.
• We don't publish the captioned video to your channels: we hand it over ready, publishing stays

yours.

• The captioned video stays downloadable for 7 days; **subtitle files, being text, don't

expire while the account stays active**.

What it costs

Trial: 14 days and 60 minutes, no card, up to 60 minutes per video, with the captioned video with the echoflow watermark. Pro, the only paid plan, removes the watermark, adds style customisation and output translation, and comes in two sizes: €19/month for 200 minutes or €49 for 1,000, with up to 10 people and separate workspaces per client. The annual option is worth two months free.

Want to see what the result looks like first? The demo is a real, browsable job, with no sign-up.