Product photos. Product videos. One studio.

Try five images total and three free five-second video generations once per verified account. Paid plans include separate monthly image and video allowances: Starter has 25 images and 3 video credits; Pro has 75 and 10; Business has 250 and 30. One video credit covers five seconds; two cover ten. A US$29 top-up adds 10 non-expiring video credits to any account.

Credits are reserved when you submit and used only when a completed video is saved. Technical failures return the credits automatically. If included or bonus credits expire during a failed job, replacement credits have their own visible expiry. A completed video uses credits even if you prefer a different creative result; each new version is a new generation.

Video currently starts from a JPG, PNG or WebP image, including images made in SceneForge. Direct 3D model uploads and editing an existing video are not available. For a new version, reuse your opening image and change the motion.

Choose five or ten seconds, landscape (16:9) or portrait (9:16). Videos download as 720p MP4 files and may include generated audio. Use your editor to mute or replace audio when needed. Videos continue generating after you close the page and stay in your account library.

Monthly video credits expire at the end of their paid billing period and do not roll over. Purchased credits do not expire. The three free trial credits are a one-time grant, not a monthly allowance, and can only make five-second videos. If a Business launch bonus is offered, it adds 20 credits once, expiring 30 days after the grant. That bonus is separate from the recurring 30 credits/month.

SceneForge currently uses Google Gemini 3.1 Flash Image for image generation. It creates and refines product scenes from your reference photos and instructions, with image sizes up to 4K on eligible plans. AI can change product details, so review the result before use.

You can upload up to 14 product reference photos on eligible plans. Clear views from several angles give the model more information about shape, materials and details. Surfaces not shown in the references are generated, not verified from a 3D model, and may differ from your real product. Compare the result with your original photos before use.

Product mode takes your product photos and places them into an entirely new AI-generated scene — great for creating lifestyle images from scratch. Edit mode takes an existing scene photo (like a room) and modifies it according to your instructions — perfect for virtual staging, adding products to real spaces, or restyling existing rooms.

After generating an image, you can refine it with follow-up instructions like "move the product to the left" or "make the lighting warmer." The AI remembers the full conversation context, so each refinement builds on the previous result. You can also attach a reference image during refinement to show the AI exactly what you want — a texture, a style, or a specific product to add.

The AI Prompt Builder is a built-in assistant that helps you write effective prompts. Instead of writing one long description, you fill in structured fields — product, scene, lighting, camera angle, props, style, and exclusions — each with example presets you can click to use. Hit "Build prompt" and AI combines your inputs into an optimized prompt. Included free on all plans.

A seed controls one source of randomness. Reusing it can help compare prompts with the same model and settings, but identical results and consistent product details are not guaranteed. Image seed control is available on Starter and above. Videos use automatic seeds.

Default generates at ~1024px (fast, good for previewing). 2K generates at ~2048px (great for web and social media). 4K generates at ~4096px (print-ready quality). Higher resolutions take longer to generate. 2K is available on Free and Starter plans; 4K requires Pro or Business.

Thinking level controls how much the AI reasons before generating. Default is fast and works well for most scenes. HIGH makes the AI spend more time planning complex compositions with many elements — useful for detailed scenes with specific product placement, multiple props, and precise lighting. Available on Starter plans and above.

When enabled, the AI can include realistic human figures in your scenes — useful for lifestyle product shots showing people using your furniture or products. You can choose to allow all people, adults only, or no people. Available on Pro and Business plans.

You can choose from 10 aspect ratios: 16:9 (hero banners, landscape), 1:1 (Instagram, thumbnails), 9:16 (stories, reels), 4:3 (standard photos), 3:2 (DSLR ratio), and more. The aspect ratio affects the shape of the generated image and should match where you plan to use it.

The creativity slider (temperature) controls randomness. Low values (0-0.3) produce predictable, safe results. Medium (0.5-0.8) gives balanced output — recommended for most use cases. High (1.0+) produces more creative and varied results but may introduce artifacts. Default is 0.7.

You get 5 free image generations in total. Trial generations count toward that allowance, so signing up does not add more. Starter ($19/mo) gets 25/month, Pro ($49/mo) gets 75/month, and Business ($129/mo) gets 250/month. Paid allowances reset after each successful monthly renewal. Unused generations do not roll over.

Each time you hit "Generate" counts as one generation — whether it's a new scene or a refinement of an existing one. Failed generations (timeouts, errors) are not counted against your limit.

Each scene has a refinement cap — the maximum number of follow-up edits you can make to a single generated image. Free plans allow 1 refinement per scene, Starter allows 2, Pro allows 4, and Business allows 5. Starting a new generation resets the count.

Yes. All images you generate with SceneForge are yours to use for any commercial purpose — product listings, marketing materials, social media, print ads, etc. There are no watermarks on any plan.

Generated images are delivered as high-quality PNG files. You can download them directly from the app with one click.

Most generations complete in 15-60 seconds depending on resolution and complexity. 4K images with HIGH thinking enabled can take up to 2-3 minutes. If a generation times out, try reducing resolution to 2K, thinking to Default, or start a new conversation to shorten the context history.

You can upload PNG, JPEG, and WebP images as product references or scene images. For best results, use well-lit photos with clean backgrounds and consistent lighting across multiple angles.