Snap has spun off its advanced generative AI video research team into an independent gaming-focused venture named Dotmo to offload high compute costs.
The staggering financial reality of maintaining cutting-edge machine learning infrastructure is forcing major social media platforms to drastically rethink their product roadmaps. In an aggressive corporate restructuring effort to pull heavy research and development liabilities off its balance sheet, Snap Inc. has announced the spinoff of its internal generative artificial intelligence video engineering group into a completely separate startup named Dotmo.
The parent company behind Snapchat cited the high capital expenditure required to conduct advanced generative video R&D inside its core business as a primary catalyst for the separation. The newly independent entity, Dotmo, will pivot away from purely social messaging filters and instead focus its engineering resources on building foundational AI models optimized for interactive gaming environments and 4D video diffusion.
Snap Financial Blueprint: [ML & AI Compute Overhead] ──> ~$100M Expended Per Quarter Corporate Restructuring: [Spinoff of Advanced Video R&D] ──> Shifting Capex into 'Dotmo'
(Converts Headcount & Cloud Drag into an Equity Holding)
The Unpacking of a Pragmatic Spinoff Model
The structural layout of the deal allows Snap to offload intense computational bills while retaining a clear path to commercial upside. Under the framework of the spinoff, the initial core team at Dotmo will consist entirely of outgoing Snap engineers and researchers who are leaving the social media company to launch the venture.
In exchange for this transfer of talent and an exclusive technology license covering Snap’s early mobile-device video research, Snap will receive a substantial equity stake in Dotmo. Furthermore, Snap will license its core software back from the startup for integration into future gaming and augmented reality (AR) features inside Snapchat down the line.
The transaction features a unique personal element. Snap’s co-founder and Chief Technology Officer, Bobby Murphy, is acting as Dotmo’s lead angel investor, putting his own capital behind the seed round. Murphy will retain his full-time executive role at Snap, leading its broader on-device machine learning roadmap, while holding a major personal equity position in Dotmo.
While this governance structure gives the new venture immediate validation, it will require close board oversight to manage potential overlapping interests regarding talent allocation and technology transfer priorities between the two connected entities.
Balancing the Cloud Ledger
The creation of Dotmo is a direct reaction to the rising costs of competing in the frontier model race. Public financial declarations show that Snap has consistently spent roughly $100 million per quarter on machine learning and compute infrastructure to power its current user recommendations and basic AR lenses.
This is the second time Snap has used an isolation model to shield its primary balance sheet from cash-burning experimental tech. Earlier this year, the firm established a distinct corporate subsidiary for “Specs,” its augmented reality smart glasses hardware line, to separate physical product manufacturing hazards from its core advertising operations.
By establishing Dotmo as a standalone enterprise capable of raising outside venture capital, Snap successfully converts fixed infrastructure overhead into a flexible asset. The move reflects a broader trend across the tech sector, where mid-tier social applications are finding that while building custom generative video models is a fascinating research pursuit, running those models daily is an expensive luxury best financed by outside capital markets.
