OmniHuman 2.0 pricing guide<\/a> explains how to size a project before you hit generate.<\/p>\nThis is the uncomfortable bit for a short listing reel: if you need only eight seconds of talking agent, don’t assume you’ll be charged for eight seconds on a tier whose listed output starts at 30. Budget for an eligible generation. Check billing in your account. BytePlus also lists an avatar tier at $0.12 per second, but the product page doesn’t specify the same 30-second range for that tier, so don’t copy the film-grade minimum onto it.<\/p>\n
My preference is to script a useful 30-second host segment if the film-grade tier is the plan, then cut it into the final listing edit. Paying to generate a half-minute just to keep one generic greeting is a poor trade. Save a shot list first and make the agent deliver something the room footage can’t: context, a clear name and a reason to request an in-person viewing.<\/p>\n
Edit with a property-accuracy checklist<\/h2>\n
Before export, play the video with the listing record open. Pause on every number and every room claim. Check the price, address, availability, square footage and the footage rights against the approved materials, then remove anything that can’t be supported. If the listing changes, re-record the affected line or replace the dated caption; don’t leave a previously correct price circulating as if it were current.<\/p>\n
Make the visual handoff obvious. Make a clean cut. A blend can imply a continuous tour that never happened. Label the presenter segment as AI-generated where your publishing context calls for it, and don’t pass a generated room background off as an image of the property. Keep a copy of the approved script, the property media permissions and the final export together. When a colleague asks which version went out, you should be able to answer without hunting through chat attachments.<\/p>\n","protected":false},"excerpt":{"rendered":"
Make an AI real estate listing video from a photo with a short talking-agent intro, real property footage, verified claims and clear cost math.<\/p>\n","protected":false},"author":2,"featured_media":111,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[],"class_list":["post-109","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-use-cases"],"markdown":"An AI real estate listing video from a photo works best as a short agent introduction, not a fake walk-through. BytePlus says OmniHuman 2.0 animates one image with audio and an optional text prompt; put the real property footage after the talking segment, and check every spoken claim against the current listing before you publish.\n\n## Keep the property real\n\nThe easiest mistake is asking a portrait-to-video tool to show a kitchen it has never seen. A headshot can't verify the counters, the room size or the view from the balcony. Treat the generated segment as the host of the listing, then cut to actual footage or licensed listing photos for the rooms. Keep the two shots distinct. If a viewer asks where a feature appears, you should be able to point to the source photo or an approved property record.\n\nBytePlus describes OmniHuman 2.0 as generating video from a single image and a voice track, with an optional text prompt for character actions and camera movement. It does not describe the model as a property-tour engine that checks listing details. That distinction matters. Animated agent first, documented home second. For a more general starting point, see [how to use OmniHuman 2.0](\/blog\/how-to-use-omnihuman-2).\n\nGet the agent's permission to animate their photo and voice. If the agent isn't the speaker, don't make the video imply they recorded the narration. Ask your broker or legal team how to label a synthetic presenter under the rules where the listing will run; this article isn't a statement about local advertising requirements. Say what you generated. The property footage must remain an honest depiction.\n\n## Build a 45-second listing script\n\nWrite the copy from the verified listing sheet, not from a property portal's old cache. A tight structure makes it easier to replace a single changed fact without rebuilding the whole edit:\n\n1. **Agent introduction, roughly 8 seconds.** State the agent's name and identify the listing without calling it a live tour. Keep the agent on screen here.\n2. **Property footage, roughly 25 seconds.** Show real images of the entry, living area and any feature the narration actually mentions. Use captions for figures such as asking price or floor area only after checking the current approved listing.\n3. **Next step, roughly 12 seconds.** Show the agent again only if you have a second approved clip, or finish over property footage with the actual contact route and a date-aware call to action.\n\nThose timings are an editing example, not a BytePlus performance guarantee. Read the first eight seconds aloud. Names trip up rushed scripts. If the name, address or price turns into a breathless list, cut the adjectives before you speed up the voice. You can always keep the property footage silent under the intro and give the room shots their own recorded narration later.\n\nA safe sample opener, with placeholders left deliberately blank, is: \u201cI'm [approved agent name]. Here's a quick look at [approved listing address]. The next shots show the actual property; contact me for the current details.\u201d Replace the brackets from an authorized listing record. Don't make up neighborhood claims. Good copy doesn't need them.\n\n## Prepare the presenter image and audio\n\nUse a clean chest-up image of the consenting presenter, facing near the lens. Keep their mouth unobstructed, leave room around the shoulders and avoid a hand cropped halfway through the frame. The home belongs in the next shot. If the only available photo has a busy office backdrop, a plain authorized portrait is the better source; see the [reference-image guide](\/blog\/omnihuman-2-reference-image-tips) for framing decisions.\n\nRecord the agent's own voice if that's the intended identity. Listen for clipped street names and swallowed numbers before you submit an audio file. Almost right is still wrong. Cut long pauses in your editor, but don't remove the space needed to pronounce the street or the unit number cleanly. See the [lip-sync guide](\/blog\/omnihuman-2-lip-sync) for the parts of the audio and portrait that affect a talking-face shot.\n\nKeep the motion request modest: \u201cPresenter faces camera, speaks calmly, makes a small open-hand gesture once, and stays in a stable waist-up frame.\u201d That's an example prompt, not a tested output. BytePlus says text prompts can control gestures and camera movement; it doesn't promise any particular prompt will produce a specific gesture on every run. Skip the sweeping pan. Skip invented furniture. Let the actual room footage carry the real-estate detail.\n\nIf you want to assemble a first version, [make a video on the OmniHuman2.net overview](https:\/\/omnihuman2.net) and then bring the resulting presenter segment into your normal editor. The site is an independent guide, not an official BytePlus service; check the creation route and current access terms before uploading a client's media.\n\n## Budget for generation, not just the final cut\n\nBytePlus lists the film-grade tier at $0.07 per generated second and describes its video output as 30 to 300 seconds. A 45-second generation is **45 s \u00d7 $0.07 = $3.15** at that listed film-grade rate. If you generate two alternatives, the generation portion becomes **2 \u00d7 45 s \u00d7 $0.07 = $6.30**, assuming both are billed at that rate. Editing time and any other service charges are outside that math. The [OmniHuman 2.0 pricing guide](\/blog\/omnihuman-2-pricing-video-cost) explains how to size a project before you hit generate.\n\nThis is the uncomfortable bit for a short listing reel: if you need only eight seconds of talking agent, don't assume you'll be charged for eight seconds on a tier whose listed output starts at 30. Budget for an eligible generation. Check billing in your account. BytePlus also lists an avatar tier at $0.12 per second, but the product page doesn't specify the same 30-second range for that tier, so don't copy the film-grade minimum onto it.\n\nMy preference is to script a useful 30-second host segment if the film-grade tier is the plan, then cut it into the final listing edit. Paying to generate a half-minute just to keep one generic greeting is a poor trade. Save a shot list first and make the agent deliver something the room footage can't: context, a clear name and a reason to request an in-person viewing.\n\n## Edit with a property-accuracy checklist\n\nBefore export, play the video with the listing record open. Pause on every number and every room claim. Check the price, address, availability, square footage and the footage rights against the approved materials, then remove anything that can't be supported. If the listing changes, re-record the affected line or replace the dated caption; don't leave a previously correct price circulating as if it were current.\n\nMake the visual handoff obvious. Make a clean cut. A blend can imply a continuous tour that never happened. Label the presenter segment as AI-generated where your publishing context calls for it, and don't pass a generated room background off as an image of the property. Keep a copy of the approved script, the property media permissions and the final export together. When a colleague asks which version went out, you should be able to answer without hunting through chat attachments.\n","_links":{"self":[{"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/posts\/109","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/comments?post=109"}],"version-history":[{"count":1,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/posts\/109\/revisions"}],"predecessor-version":[{"id":110,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/posts\/109\/revisions\/110"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/media\/111"}],"wp:attachment":[{"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/media?parent=109"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/categories?post=109"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/omnihuman2.net\/blog\/wp-json\/wp\/v2\/tags?post=109"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}