Xross Road Unveils HANASEE-image-1.0 for Vertical-Scroll Manga

On April 24, 2026, Xross Road announced the release of HANASEE-image-1.0, a proprietary image generation model purpose-built for vertical-scroll manga panel generation. The model will be available through the closed beta version of the manga creation platform HANASEE, launching on April 30, 2026.

This isn't another general-purpose image generator slapped with a new name. The company explicitly states the model was developed under the supervision and collaboration of professional manga artists. That distinction matters when you consider how frustrating it is to watch generic AI tools hallucinate character faces across consecutive panels (a problem that has plagued users for years, frankly).

According to the official press release, HANASEE-image-1.0 is built on an open-source image generation model foundation but has been further refined and additionally trained specifically for manga panel generation. The core technical challenge here is sequential consistency—keeping characters looking like themselves across dozens of panels while maintaining narrative continuity.

The model addresses three primary technical requirements. First, character consistency: the system incorporates proprietary optimizations that preserve character identity including hairstyle, facial features, and clothing across panel images depicting a wide range of scenes. Second, manga-panel-specific art style with consistency: at launch, the model offers pre-optimized art styles including "sawayaka" (fresh), "fuwafuwa" (fluffy), "Bright," "Kawaii," "Powerful," and "Cool." Third, composition tailored for vertical-scroll manga: the model supports flexible composition generation that adapts to specified panel sizes and content.

Independent reporting from Animation Magazine corroborates the timeline and provides additional context about the company's funding and partnerships. Xross Road raised $1.5 million USD in pre-seed funding co-led by Arbitrum Gaming Ventures and Decima Fund, with participation from Taisu Ventures and Baboon VC. The company also disclosed partnerships with the TV Asahi Group.

CEO Yosuke Utsumi stated the platform is designed to enable professional-quality manga production while opening doors for individual creators to become architects of the next generation of IP. The system uses specialized AI agents for different stages: illustration, scripting, and panel layout. This layered, isolated editing tool is meant to overcome common consistency issues found in general image generation AI.

The physical workflow is where this gets interesting. Creators start from a novel or story seed, then the platform produces serialized manga with consistent characters, worldbuilding, and art style. The creative workflow determines optimal panel size and framing based on story progression, and the model generates compositions accordingly. This enables the rhythm unique to vertical-scroll manga storytelling to be expressed at the panel level.

Closed beta access will be granted by lottery from among those who register via the application form by April 27, 2026. The company's official website describes HANASEE as enabling users to turn ideas and stories into vertical-scroll manga through a combination of proprietary image generation models and structured creative workflow.

The broader ecosystem includes IP-LINK, an infrastructure layer integrating data across the entire IP supply chain, and XR, a foundational asset supporting activities within the Xross Road ecosystem. Token holders get economic stake in value generated through Xross Road's anime-related projects, plus governance rights for key decision-making processes.

Whether this actually solves the character consistency problem that has frustrated manga creators for years remains to be seen. The closed beta lottery closes in three days, and the real test begins when actual creators start trying to produce coherent serialized work without spending hours manually correcting AI hallucinations. Time will tell if the proprietary optimizations deliver on the promise.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn