Skip to main content

BULIN 47 AI GUIDE

How to make a Bulin 47 AI video

To make a Bulin 47 AI video, sign in with Google, upload one portrait and use a generation credit. The fixed freestyle reference guides the performance; follow your task in My videos.

Before you start: one photo, one 10-second freestyle

To make a Bulin 47 AI video, bring one clear portrait and sign in with Google. The studio prepares the freestyle reference, prompt and original soundtrack, so you do not need to film yourself, upload a performance video or work out model settings. One accepted generation uses one paid credit.

Your portrait casts the main rapper originally wearing sunglasses and a pale pink shirt, behind the person holding the speaker in the foreground. It does not select another crowd member. You receive a 10-second, 720p, 9:16 MP4 with the matching reference soundtrack in My videos. Download it within 30 days and save a copy to share.

Watch the 10-second scene with sound → · Choose a pack for your plans →

What photo works best for your freestyle?

Start with a photo of the person you want to become the main rapper, rather than a screenshot of the entire crowd. A head-and-shoulders portrait is useful: the face should be large enough to inspect, but you do not need a full-body pose or a photo taken beside a speaker. A photo gives appearance information; the motion reference supplies the performance. Use an image you own or have permission to upload, with the person’s consent.

  • One person: crop a group photo so the intended face is the clear subject.
  • Even light and sharp focus: avoid dark faces, motion blur and heavy beauty filters.
  • Visible features: avoid sunglasses, hands across the face or hair covering the eyes.
  • JPG, PNG or WebP, up to 10 MB. Each side must be 301–5760 pixels; width divided by height must be greater than 0.4 and less than 2.5.

There is no required fixed photo ratio such as 9:16. The upload limits above exclude excessively narrow or wide images, while the video’s portrait format is a separate output choice. Leave enough room around the head to avoid cutting off important features. Renaming an unsupported file to .jpg does not convert it; export a real supported image if your original is HEIC or another format.

Follow the studio workflow

Prepare a photo and browse the examples before starting. The studio handles the fixed reference, so you can focus on your portrait and task. There is no need to download an app, learn the choreography or choose a custom model.

  1. Prepare the portrait using the checklist, then open the studio.
  2. Sign in with Google. Use the same account later to access credits and private tasks.
  3. Select exactly one photo. Check the preview and replace the photo before submission if it shows the wrong person.
  4. Check your credit balance and choose a pack if needed; each accepted generation uses one credit.
  5. Submit once and follow the task in My videos. Leaving the page does not cancel an accepted task; check its existing status before submitting again.
  6. Review the completed clip, download the MP4 and save a copy before its 30-day storage period ends.

Understand the reference and preview formats

The fixed motion reference uses the first 10 seconds of the supplied freestyle clip and retains its original audio. It guides the rapper’s movement, rhythm, camera and surrounding people. Your download is a 10-second, 720p, vertical 9:16 MP4. Review your finished video for identity consistency and audio alignment instead of judging the reference alone.

The 10-second reference guides the main rapper’s movement, rhythm, camera and the surrounding crowd. Its soundtrack is added to the generated video.

The six gallery examples are creator-supplied clips, each 6 seconds long, 360 × 640 and without sound. They show the visual idea and include creator credits. Use the separate motion reference above to preview the 10-second scene and its soundtrack; your own result appears privately in My videos.

Review the whole result before sharing

Watch beyond the opening frame. Check the face as the rapper turns, the hands as they pass the body or speaker, and the shape of clothing during movement. Look at the surrounding crowd for unexpected changes and the camera for jumps or stretching. Listen for whether the audio starts and ends cleanly, then compare visible gestures and mouth movement with the rhythm. A familiar beat alone does not prove precise lip synchronization.

How to share your Bulin 47 AI video on TikTok

  1. Open My videos with the Google account you used to generate. When the task is ready, download the MP4 to your phone or computer within 30 days.
  2. For captions or finishing edits, open the downloaded file in an editor such as CapCut. Keep the vertical 9:16 canvas and check that the reference soundtrack is not muted or replaced unintentionally.
  3. Upload the saved clip to TikTok or your chosen platform. Preview the whole video and audio, label it as AI-generated and check the platform’s current sharing and music rules. The studio downloads a file for you; it does not publish to your social account.

Troubleshoot without duplicate submissions

  • Blurry face or wrong person: inspect the source portrait and crop first. A sharper, single-person photo may be a better next input; there is no guaranteed repair or face-selection control for an existing result.
  • No sound: the six gallery examples intentionally have no audio. For the reference or your result, check player mute, device volume and the downloaded file before reporting a missing audio track.
  • Stretched video: preserve the MP4’s aspect ratio in your player or editor. Fitting a portrait clip into a different canvas by stretching can distort faces; use proportional scaling instead.
  • Canceled payment or uncertain submission: check the receipt, credits and My videos before paying or submitting again. An accepted task may still be processing; check My videos before starting another version. If the state remains unclear, contact support with the task or receipt details.

Know what a retry or download means

A confirmed failed task returns one generation credit. A request under support review keeps the credit reserved until its outcome is resolved; it is different from a confirmed failure. A completed video whose appearance you dislike is also a different situation: a new generation normally uses another credit, and credit restoration is not the same as a cash refund. Read the refund policy or contact support instead of assuming that deleting a video refunds it.

Uploads are private to your account. Photos are kept for up to 7 days and finished videos for 30 days, so save the MP4 locally. When sharing, identify it as AI-generated, respect the subject’s consent and check the platform’s rules for your intended use of the footage and audio. The site does not automatically post to TikTok or provide a universal publishing license.

Ready to make your version?

Compare the prices and examples before planning a generation. Read the privacy and refund rules, or return to your private history with the same Google account.

Open the studio
What is Bulin 47 AI? →