Founded in 2024 by Lina Colucci, Sidney Primas, and Andrew Weitz, Lemon Slice is building a 20-billion-parameter diffusion model, Lemon Slice-2, that produces video avatars from a single image and streams them at 20 frames per second on a single GPU. Delivered through an API and a one-line embeddable widget, the technology lets customers adjust backgrounds, styling, and character appearance while drawing on ElevenLabs for voice synthesis. The avatars can be human or non-human and are designed to sit on top of knowledge bases so they can address customer support, tutoring, language learning, e-commerce assistance, or mental-health scenarios. The company emphasizes safety with guardrails against unauthorized face or voice cloning and uses large language models for content moderation. Lemon Slice differentiates itself from rivals such as D-ID, HeyGen, and Synthesia by training a general-purpose video diffusion transformer that, according to investors, can eventually cross the uncanny valley. The eight-person team plans to expand its engineering and go-to-market staff and cover compute costs for further model training with newly raised capital. No revenue or user figures were disclosed in the article.

Total raised

$11M

Funding rounds

1

Latest round

Seed

Latest activity

Dec 2025