San Francisco startup The Biological Computing Co. has released TBC Mischief 1, an image-to-video model that turns still images into short clips. The company claims insights taken from living neurons make generation about five times faster than the base model at the same quality. The price is set at $0.005 per second of video, which matters for businesses where large volumes of clips determine production costs.
Mischief 1 pricing, availability and speed claim
TBC Mischief 1 is built on Wan 2.2, an open video model from Alibaba, and is distributed through the developer platforms fal and OpenRouter. The listed price means a five-second clip costs 2.5 cents, a figure the company disclosed at launch on Thursday. Positioning on established developer platforms lowers integration friction because teams can call the model through familiar inference channels rather than deploying separate infrastructure.
The claimed acceleration does not come from new graphics hardware or a larger model. TBC keeps living neuronal networks in its own laboratory and studies how they respond to signals. The findings are then converted into a small software add-on that runs on normal AI chips. That design removes the need for customers to buy wetware or specialized equipment to use the result.
The performance statement remains a company claim without outside confirmation. TBC has not published independent benchmarks supporting the fivefold speedup at equal quality. For buyers, the absence of third-party tests means speed, visual stability and consistency across different image types still need direct evaluation before committing production workloads.
What neuron-derived acceleration means for business video
For companies producing marketing clips, product demonstrations and training visuals, cheaper and faster image-to-video generation changes unit economics. Lower cost per second makes short-form content, A/B variants and localized versions easier to produce in volume. Small firms gain access to video automation without rendering farms, while large firms can scale output across catalogs and regional campaigns.
The main limitation is verification of quality at speed. No independent data shows how the model handles faces, text, complex motion or brand-specific assets. Buyers should test output on their own images, measure actual latency on fal and OpenRouter, and clarify usage terms, support and data handling. This release alone does not prove that biology-derived optimization works across other model types.
The marker to watch is the commercial rollout through the September deal with Amazon Web Services to sell TBC models to businesses. Adoption volumes, published customer cases and independent speed measurements will show whether the approach lowers costs in practice. A separate path also bears watching in Singapore, where a rack of living neurons is already running to perform computing directly.
