NewsPublished:
TBC mischief 1 Is a Video Model Sped Up by Lessons From Living Neurons
The Biological Computing Co. says its Wan 2.2-based image-to-video model runs 5x faster at half a cent per second, now on fal and OpenRouter.

Most AI speed-ups come from the same toolbox: quantization, distillation, caching. The Biological Computing Co. (TBC) says its new video model got faster from somewhere much stranger: measurements of living neurons grown in a lab. Its first commercial product, TBC mischief 1, launched on October 30, timed to Mischief Night, and it's available right now on fal and OpenRouter.
What mischief 1 actually is
- An image-to-video model built on Wan 2.2, the widely used open-weight video model
- TBC says it matches Wan 2.2's quality while generating video about 5x faster
- Priced at half a cent per second of video, so a five-second clip costs 2.5 cents; TBC says that makes it the cheapest image-to-video option on both marketplaces
- The company's add-on software adds less than 0.1% to the base model's parameter count and runs on ordinary AI hardware
Where the neurons come in
According to TBC, its lab grows neurons, measures how they process information, and turns the computational principles it finds into software that makes existing models more efficient. The neurons themselves stay in the lab. Nobody is running a video model on a dish of cells; what ships is regular code inspired by, and in TBC's words "derived from", biology.
It's worth being clear that the speed and quality claims are the company's own, and TBC hasn't published the details of what exactly it borrowed from the neurons. One AI studio, Wonder Studios, says it tested mischief 1 in its production workflow and liked how much more iteration it could fit into the same budget.
Why I find this interesting
For anyone who makes fan animations, storyboards or quick animatics, the price is the headline: cheap enough to throw dozens of takes at a single shot. The bigger question is whether the biology angle holds up. TBC says it's already applying the same process to newer video models and working on language and robotics models next, so this is a story I'll be watching beyond a single launch.
Source: TBC press release on PR Newswire
Source: prnewswire.com