Back to home

Sheru Bhai AI User Guide

Learn how to prepare photos, choose scenes and casting, generate images and turn a selected image into a video.

Last updated: 2026-11-16

Start with one companion photo

Open the editor. The starting combination is Image → Add Sheru to my photo. Choose the result you want, then prepare the photo. On a desktop, settings sit beside the preview; on a narrow screen, they follow in a single column.

The AI models behind your creations

Image generation and photo editing use GPT Image. Video generation uses Wan. When a video requires a new starting image, GPT Image creates that image before Wan generates the video.

1. Choose your photo composition or scene

Who is in the scene? offers four choices. Add Sheru to my photo is selected for a new creation. Choose Sheru’s pose or interaction above the photo upload: stand, sit, wave, arm around shoulder, toast or gift. This selection is applied during generation; the preview continues to show your original photo until then. Upload an individual or group photo to keep its people, clothing and background while adding Sheru. Use additional direction to describe placement and interaction, such as “Have Sheru sit on my left.” The preview shows your original photo before generation and the edited image afterward. The edited image can also continue into video.

Choose Me & Sheru to select one of the six scenes below and apply your photo identity to that composition. Become Sheru and Only Sheru also use this scene library.

Image and video share the same scene library. Read the image composition and video action before choosing:

  • Street companions: a clear street portrait; the video continues the walk, adds a brief interaction and finishes facing the camera.
  • Grand arrival: stand beside a parked luxury car; the video uses a small cuff adjustment and a look toward the camera, with the car and doors still.
  • Golden throne: stand beside the throne in a palace; the camera moves closer and the subject offers a small nod.
  • Billionaire bids: an auction portrait with a raised hand and a rival; choose the car or palace as the auction item, then build the reaction around that scene.
  • Private jet: stand inside the cabin beside a window, holding a glass; the video turns toward the camera for a restrained toast.
  • Birthday surprise: stand beside a table with the cake already in place; add the recipient and greeting, then continue with a gesture toward the cake and a confetti portrait.

The preview helps you understand composition and action. Keep additional directions consistent with the selected scene, especially the position of hands and props. See the scene and prompt examples for starting ideas you can edit.

2. Choose who appears

  • Add Sheru to my photo: keep your original people and environment, and choose Sheru’s pose and interaction.
  • Me & Sheru: use a scene template to appear alongside him.
  • Become Sheru: use your build and pose as a reference for a lion transformation. This changes the face into a lion character; it is not a human-face companion photo.
  • Only Sheru: make a Sheru character image or story without a personal photo. A scene can still contain supporting fictional characters, such as an auction rival.

Changing the casting choice changes the photo requirement. Choose the cast before uploading so the reference matches your intended result.

3. Prepare your photo

Use a clear, well-lit photo. Add Sheru to my photo supports individual and group photos; available space beside the people helps create a natural composition. For a scene-template portrait, keep the face visible and avoid heavy filters or obstructions. For a transformation, use a readable standing pose with enough room around the shoulders, hands and body. Avoid collages, screenshots with interface elements and very small faces.

The editor accepts JPEG, PNG and WebP, up to 10 MiB per file, with the longest edge no greater than 4096 pixels. Resize a larger photo before selecting it. Input dimensions describe the uploaded photo; the output resolution and aspect ratio are separate settings.

Only use photos you have permission to process. Selecting a photo creates a local preview; submitting real generation uploads the required files. Refreshing the page can require you to select the photo again. Removing it from the form does not delete an already uploaded copy. See the Privacy Policy or request deletion.

4. Set the output and review the quote

Open More settings to choose image aspect ratio, resolution and quantity, or the available video settings. Use portrait framing for vertical posts, square framing for a compact portrait, or landscape framing when the scene needs more space. Image and video settings are saved separately while scene, cast, photo and additional direction remain shared.

For images, choose the available resolution and number of versions. A larger resolution adds output detail; it does not change which person appears. More versions let you compare compositions and change the displayed credit cost.

The standard video uses 720p, 15 seconds and music/sound effects. Set the aspect ratio to suit where you plan to share it. A scene's story description is a creative direction, not a promise of spoken dialogue. Use the actual sound option shown for your request when reviewing the output.

Use the additional prompt for a few concrete details: lighting, expression, camera movement or a prop to preserve. For example: “Warm evening light, keep both faces visible, a slow camera push-in.” Avoid conflicting instructions such as asking a standing subject to be seated in the same frame.

Review the credit quote before submitting. A new-frame video includes image creation and video creation; continuing from an existing image uses that selected first frame. Pack prices and credit entitlements are listed on Pricing. A real request requires sign-in and sufficient credits.

5. Generate and choose a result

Submit once and follow the task status. Generation continues asynchronously. If the connection is interrupted or submission is uncertain, use the original task's recovery action instead of repeatedly submitting a new request.

When images arrive, compare the returned versions and select the one you want. Check likeness, hands, clothing, framing and important objects before continuing. Save the files you want to retain while they are available.

6. Make a selected image move

Choose a real image result and click Animate this image. The editor changes to video and binds that image as the starting frame, keeping the scene and subjects. Review the video settings and quote before submitting.

Changing the scene, cast or character identity asks you to release the first-frame binding. Keep it if you want to preserve the selected composition; release it to create a different scene. Newly selected photos must meet the same upload requirements. After a page refresh, reselect a result from history or supply the required photo again.

History, downloads and common questions

Where is my task? Sign in to the same account and open creation history in the editor. Text and settings saved in the browser tab are a draft, not proof that a request was submitted. Account task history is the place to check a real request.

The task is still processing or needs recovery. Query or recover the original task using its available action. Do not equate a temporary missing result with a confirmed failure. A newly submitted generation is a separate request and charge.

Only some results arrived. Keep the returned files and check the task status. Partial output is distinct from complete success or complete failure. Contact support with the task identifier if the result remains incomplete. Confirmed failed stages restore their deducted credits once; completed stages remain charged.

A download link expired. Use the task's refresh action while the file is still retained. Refreshing a signed link does not renew the file's storage period. Provider-hosted outputs are temporary, with up to 72 hours of retention; use the expiry shown for the result and download promptly.

The generation button asks for a change. Follow the displayed message: sign in, supply the required photo, select a supported option, recover a pending request or add credits as appropriate. Changing the draft does not cancel a running task.

For help, use Contact & support. Include the task identifier and the displayed status without sending passwords, payment credentials or private media links. Review the Terms of Service before uploading or sharing your work.