Perplexity confirmed Wednesday that it has brought the core team from Visual Electric into the company, in a deal aimed at accelerating Perplexity’s work on visual content generation and multimodal search. Financial terms were not disclosed. Sequoia Capital had been an investor in Visual Electric, a San Francisco startup known for tools that let users generate images and, more recently, short videos from text prompts.

The integration follows months in which Visual Electric expanded its product beyond static images to support video generation, a technically demanding addition that requires both novel model architectures and significant compute resources. Perplexity, which built its early reputation on natural‑language search using large language models, said the acquisition would help it add richer visual answers and generative media to its product roadmap.

“We see visual generation as a natural next step for search,” Perplexity’s chief executive said in a statement. “Bringing Visual Electric’s team into Perplexity will let us move faster on building safe, grounded multimodal experiences.” A Perplexity spokesperson declined to provide timing for new features or to disclose whether Visual Electric’s former codebase and models will be integrated wholesale into Perplexity’s systems.

Industry analysts said the deal fits a broader pattern of consolidation in the AI ecosystem, in which emerging startups with specialized generative capabilities are being absorbed by larger platforms seeking to offer end‑to‑end services. “It’s an acqui‑hire that gives Perplexity not just talent but also domain expertise in visual synthesis,” said Marina Lopez, a researcher at the Center for Digital Innovation. “But it also raises immediate questions about content control and verification.”

Those concerns are acute for video generation, which can enable highly realistic fabricated footage. Companies building generative media face pressure from regulators, platforms and advertisers to implement robust provenance, watermarking and moderation. Perplexity said it plans to apply the same safety and grounding standards it uses for text generation to visual outputs, but provided few technical specifics. “Safety is nonnegotiable,” the Perplexity statement said, adding that the company is investing in automated filters and human review workflows.

The acquisition also underscores the escalating infrastructure demands of multimodal AI. Generating plausible video at scale requires cloud GPU capacity and specialized orchestration. That reality has drawn more attention to the role of cloud providers, whose pricing and availability can shape which companies can deploy these features economically. Competitors from large technology firms to startups are racing to offer comparable multimodal search and creation tools, turning talent acquisitions into a fast route to parity.

For Visual Electric’s team, the move offers access to Perplexity’s user base and distribution channels. For Perplexity, it is a bet that generative visuals will become a central part of search experiences rather than a peripheral novelty. How the company balances innovation with the technical and ethical burdens of video generation will be a test both for its product and for the wider market as generative AI continues to mature.