How to Turn a TikTok or YouTube Travel Video Into a Real 7-Day Itinerary

If you're like most travel-content viewers, your camera roll has a hundred saved TikToks of places you swear you'll go to. Maybe a dozen Instagram Reels of sunset cliffside bars. A YouTube Short of a market in Marrakech you bookmarked last August. The content is beautiful and the inspiration is real. The gap between 'I saved this' and 'I went there' is enormous, and the math is unforgiving: every saved video adds about thirty seconds of intent and removes about ten minutes of follow-through. The trip stays unbuilt.

Jettova's video-import flow exists to close that gap in one paste. You paste a public TikTok URL or YouTube URL (Shorts or under-10-minute videos) into the From Video tab on the iOS or Android app, or visit `/plan-from-tiktok` on the web. About ten seconds later you have a destination, a 7-day itinerary with morning / lunch / afternoon / dinner slots at real-feeling named places, a photo strip showing the five frames the AI actually looked at from the video, and pre-filled booking links to flights, hotels, activities, and an eSIM for the destination country. The 'I went there' button is a step closer than it was in the previous paragraph.

**Step one: paste a URL.** Anything public works. TikToks (`tiktok.com/...`, `vm.tiktok.com/...`, mobile share URLs). YouTube Shorts (`youtube.com/shorts/...`). Regular YouTube videos up to ten minutes (`youtube.com/watch?v=...`, `youtu.be/...`). The URL detector handles all the common forms. Instagram Reels aren't supported yet because Instagram aggressively blocks third-party scrapers, but we're working on it; for now, if you find a Reel you love, screenshot the location tag and use Jettova's normal Plan flow with that destination.

**Step two: watch the staged status messages.** This is the part where most AI products show you a spinner for thirty seconds and you wonder if it crashed. Jettova's loading screen rotates through four real status messages keyed to the actual server stages, with an elapsed-seconds counter: 'Pulling the video and caption' (0-4s, fetching the source), 'Sampling 5 key frames' (4-12s, evenly-spaced moments across the duration), 'Reading the destination clues' (12-22s, the vision model studying frames + caption), 'Drafting your 7-day itinerary' (22-30s, slotting morning / lunch / afternoon / dinner). If it goes past 30 seconds — usually a long YouTube video or a slow TikTok proxy — the message becomes 'Taking a bit longer than usual.' You always know what's happening, and you always know how long has elapsed.

**Step three: read the result page.** The hero shows the destination and a vibe line. Below it is the photo strip — the video's cover plus the five frames the AI extracted, in a 6-up grid. This is the audit trail; the visible proof that the model actually looked at your video rather than just guessing from the caption. If the destination is right, the photo strip explains why. If it's wrong, the photo strip usually explains why (sometimes a TikTok is filmed in one city but uses a different city's footage as B-roll; the strip will show that mismatch). Below the strip is the 7-day itinerary in the same format as a Jettova-planned trip, plus a row of booking links: Flights (Expedia), Hotels (Booking.com), Car rental (Expedia), eSIM (Airalo), Activities (Viator).

**Why it beats screenshotting + ChatGPT.** The obvious alternative is: take the caption, paste it into ChatGPT with 'plan me a 7-day trip,' and read the result. Two problems. First, ChatGPT can't see the video. It reads only the caption, and captions on travel TikTok are aspirational mood-text ('endless summer ✨'), not destination data. The frames are the actual location signal — the architecture in the background, the cuisine on the plate, the language on the signs — and missing them is the difference between Aruba and Costa Rica. Second, ChatGPT has no destination grounding. It'll invent a plausible itinerary that sounds reasonable but doesn't anchor to real named places in the right city. Jettova's itinerary schema forces real-feeling named places per slot — no 'a local café' filler — and we constrain the model so days 1-3 must reference things visible in the frames specifically.

**Why it beats Pinterest-tag manual planning.** The other alternative is: see the video, jot the destination, build the trip yourself across Skyscanner + Booking.com + Reddit + a half-baked spreadsheet. That works, eventually, after a Saturday. The friction is what kills it — every additional click between 'I saw this beautiful video' and 'I have a confirmed itinerary' is a chance for the trip to fall through. The Jettova flow collapses the planning step to ten seconds and the decision step to 'do I like this itinerary or do I want to swap a few activities.' Swapping is one tap. Booking is one tap per provider. The trip is built in under five minutes from a single saved video.

**What kinds of videos work best.** Short videos with clear visual storytelling. A 30-second walking-tour of a market. A 60-second clip of a hostel pool and the bar across the street. A 90-second day-in-the-life vlog. The AI samples five frames evenly across the duration, so longer videos with varied scenes give it more signal than short videos with one fixed shot. What works less well: heavily-edited videos with a lot of text overlays (the AI sees the overlay text and sometimes weights it over the underlying frame), aspirational compilations that splice together footage from five different cities (the AI may pick a destination that splits the difference), and videos where the location is filmed at night in low light (frames carry less destination signal in the dark).

**A pro move: use multiple videos for the same destination.** If you have three TikToks of Lisbon you love — one for Alfama at night, one for a Time Out Market food run, one for the Belém tower — paste each one separately. You'll get three slightly different 7-day Lisbon itineraries (the model anchors days 1-3 to the specific frames of each video) and you can pick the activities you like from each into a final plan via the normal saved-trip editor. This 'composite itinerary from three videos' approach is the closest thing to having a friend who's been to the place tell you what to actually do, because in aggregate the three TikToks are functionally one friend's recommendations.

**The rate limit, briefly.** Three imports per hour for non-admin users. The cap is there because each import costs around half a cent in vision tokens plus a few cents of bandwidth — three per hour caps the unauthorized-abuse case at a few dollars a day before we'd notice, while staying loose enough not to throttle real use. The 'I have five candidate videos for next month's trip and want to see what each one becomes' workflow fits in an hour with one or two carry-overs. If you bump the limit, just wait an hour; the import flow doesn't queue, it tells you the limit and reminds you the trip-planning funnel works without it.

**Cost: free, with the same partner-affiliate model that funds the rest of Jettova.** Pasting a URL costs nothing. The itinerary costs nothing. Saving the trip and editing it costs nothing. If you click through to one of the booking links and complete a booking on the partner site, we may earn a commission on the booking — same affiliate model as every other Jettova surface. We never see your payment information; we don't hold inventory; the partner runs the transaction. The video-import flow is a feature, not a paywall.

**To try it right now:** open the iOS or Android app and tap the From Video tab, or visit jettova.com/plan-from-tiktok on the web. Paste any TikTok or YouTube URL you've saved. Ten seconds later you have an actual trip. Whether you take it is up to you, but at least the gap between 'I saved this' and 'I went there' is no longer a Saturday's worth of planning.