Last week I helped Ah-Lin film a product intro — third sentence in, I froze, whole take wasted

Ah-Lin makes handmade earrings in Hangzhou. She usually posts on WeChat Moments and Xiaohongshu (Little Red Book) with photos and text, but the second a camera turns on, she tenses up. Last week she asked me to help record a new product intro. I stumbled on the third sentence and the whole take was dead. She asked: is there any way to make videos without showing your face? I'd just heard about LTX-2.5.

What LTX-2.5 is — and who's already shipping with it daily

LTX-2.5 is an AI video generation tool from Lightricks. You write a text description (the industry calls this a "prompt"), and it spits out clips ranging from a few seconds to over ten seconds. Ah-Lin and I dug in for a week. In the English-speaking YouTube scene, a lot of creators already use it for B-roll (background footage). For example, the tech reviewer channel "Lao Wang Unboxes" used a similar tool last week to generate five product-rotation clips. He said what used to need a photographer now takes him 10 minutes. But I made a rookie mistake on my first try: I wrote "a girl walking on the street" — the face came out mangled. Later I learned prompts need specific actions, lighting, and camera angles. Like: "25-year-old Asian woman, crossbody bag, backlit at dusk, phone-follow shot."

Today, the cost to replicate: money, time, barrier, first step

- Money: LTX-2.5 has a free quota — a few clips a day for testing. The paid plan runs about $20–30/month (~140–220 RMB, pricing may shift, check the official site). For a one-person shop, that's a coffee a week. - Time: 1–2 hours to find your feet the first time. I burned a whole weekend before the prompt patterns clicked. If you just want to try one, 20 minutes is enough for your first 5-second piece. - Technical barrier: Can you type? You're qualified. No English jargon needed — I wrote prompts in Chinese and half of them ran fine; the rest I ran through a translator. - First step: Go to ltx.studio, sign up, hit the "Create" button in the middle of the page, and paste your first description into the left-side input box. The newbie trap: writing descriptions that are too abstract. Don't write "high-end video feel." Write "black background, product slides in from the left, slow motion."

Where are you at? Here's what I'd tell you

If you're just starting out (no clients yet): I'd say hold off on paying. The free quota covers about 10 tries — enough to see if the style fits you. I free-tested for two weeks before I pulled the trigger on a paid plan. If you already have 1–2 clients: Try using it for product B-roll — gives clients an extra deliverable option. Ah-Lin now ships two versions to her clients: "real-person shoot + AI supplemental footage." Clients love the combo. If you're scaling (team of 3+): Lock down a prompt template library. Write the usuals — "product rotation," "person walking through," "empty scene shot" — into SOPs. New hires can produce on day one, no more figuring it out from scratch every time. No rush to try now. Bookmark it. Pull it out next time you're stuck in front of the camera.