QuiverAI

QuiverAI has released Arrow 2, its latest family of models for turning text prompts and reference images into clean, editable SVG graphics. The company is targeting designers, developers, and product teams who work with logos, icons, technical drawings, and illustrations that need to remain fully manipulable in tools like Figma and Illustrator, rather than flattened into pixels. This launch matters because it moves the platform beyond static generation into a workflow that also covers vectorization of raster images into editable SVGs and web-ready animations for logo reveals, loading states, motion graphics, product illustrations, and animated icons . Compared with the previous release, Arrow 2 focuses on speed and craft. The team highlights cleaner geometry with fewer, more precise control points and stronger composition around spacing and alignment, which reduces the cleanup work designers typically do on AI output before shipping. Alongside the standard model, QuiverAI is introducing Arrow 2 Telos, positioned for harder briefs that benefit from additional refinement , giving teams a choice between throughput and higher-effort generations. Both are available through the QuiverAI app and API. QuiverAI was founded by Joan Rodríguez (CEO) with cofounders Haotian Zhang, Nicklas Scharpff, and Pascal Wichmann, and is building what Rodríguez describes as an AI lab and product company focused on vector design and visual code generation. The bet is that as more surfaces (UI, brand systems, diagrams) demand structured, editable output rather than raster images, a model family purpose-built for SVG becomes a useful primitive for design tools, agent workflows, and enterprise creative pipelines.