IMAGE TO VIDEO · BY TBC
A still image. A little direction. A whole new moment. Bring your photo to life with short, expressive AI video.
Make the next moment yours.
Upload a photo or try a starter image. Your generated video will appear here.
See how a starting image and a focused motion prompt become a short clip. Try the same starting points with your own direction.
Made with Mischief 1
Slow camera movement, rolling waves, and a steady focal point.
A slow cinematic push-in toward the coastal lighthouse. Gentle waves roll toward the shore, thin clouds drift across the sky, and the lighthouse stays steady. Keep the scene calm and the camera movement subtle.
Made with Mischief 1
Subtle studio movement around a clearly framed object.
A slow camera orbit around the perfume bottle. Warm studio light glides across the glass, soft shadows shift gently on the pedestal. Keep the bottle shape, cap, and proportions unchanged.
Made with Mischief 1
Gentle environmental motion and soft afternoon light.
The flowers sway gently in a light breeze. A small butterfly drifts through the garden while the camera slowly moves forward. Preserve the flower shapes and the soft afternoon light.
TBC / WAN 2.2
Mischief 1 is an image-to-video model from The Biological Computing Co. (TBC), built on Wan 2.2. Give it a starting image and describe the movement to make a short clip for an idea, a moodboard, or a first visual draft.
Its neuron-derived technology comes from research with living neurons. Video generation itself runs as software on standard AI hardware.
Start with a clear JPG, PNG, or WebP. Leave room around your subject for movement.
Specify the action and camera movement. Keep a 3- or 5-second clip focused on one moment.
Pick 480p or 720p, review the credit cost, and generate. Preview and download the finished MP4.
Choose 480p (832 × 480) or 720p (1280 × 720), with a requested length of 3 or 5 seconds. Output is landscape video at 16 fps.
Center crop preserves proportions and trims the edges; stretch fills the frame by changing proportions. Describe specific movement rather than a long sequence of events.
This is a short, silent image-to-video model. It does not offer native vertical, 1080p, dialogue, or audio generation. Complex motion and fine details can vary between takes.
Wan 2.2 is the underlying open-weight video model. TBC applies its own neuron-derived software to improve inference efficiency; this does not make the complete Mischief 1 system open source.
TBC reports generation up to 5× faster than its base model while matching its quality. This is the developer’s claim, not a benchmark measured by ImageToVideoAIFree. Actual wait time also depends on the queue and settings.
Mischief 1 uses account credits. Choose your settings and see the full credit cost before creating a video.
These are ImageToVideoAIFree credit prices, not fal’s wholesale API rates. Each new generation has its own cost; a network interruption does not create a second generation.
Confirmed generation failures return the credits for that task.
View credit packs| Resolution | Duration | Credits |
|---|---|---|
| 480p | 3 seconds | 5 |
| 480p | 5 seconds | 10 |
| 720p | 3 seconds | 10 |
| 720p | 5 seconds | 15 |
Mischief 1 is TBC’s image-to-video model, built on Wan 2.2. It turns a starting image and a motion prompt into a short landscape video.
Mischief 1 requires sign-in and account credits on this website. If your account already has enough credits, you can use them. The daily free generator on our homepage is a separate model.
On ImageToVideoAIFree, 480p costs 5 credits for 3 seconds or 10 for 5 seconds. At 720p, 3 seconds costs 10 credits and 5 seconds costs 15. The selected cost is shown before generation.
No. This image-to-video endpoint requires both a starting image and a motion prompt. Use our text-to-video tool if you want to start with words alone.
Choose 480p or 720p, and a requested duration of 3 or 5 seconds. The model renders at 16 fps; its frame grid makes the downloaded clip slightly longer than the requested duration.
The current endpoint outputs landscape video at 480p or 720p. It does not offer native 9:16 or 1080p. Cropping or upscaling afterward is a separate editing step.
The current model generates silent video. Add music, sound effects, or voiceover afterward in an editor.
It uses Wan 2.2 as its base, with TBC’s proprietary neuron-derived optimizations. TBC claims faster generation at comparable quality; actual performance varies by workload and service conditions.
No. Living neurons are used in TBC’s research process. The discoveries are translated into software that generates your video on conventional AI hardware.
Its base model, Wan 2.2, is open-weight, but TBC describes its neuron-derived software as proprietary. Do not assume the complete Mischief 1 system is available as open-source weights.
Yes. The model is offered through fal and OpenRouter. This page provides a browser interface, so you do not need to supply your own API key.
Enhanced prompt processing can rewrite your description before rendering, and that rewrite is not controlled by the seed. Use original prompt processing, a fixed seed, and identical image and settings for more reproducible comparisons.
