Animegent

An AI creative agent that generates anime comics and storyboard videos through conversation from character design to final cut, all in one anime making studio with zero skills needed.

Comment

Inspiration

Every community has stories worth telling. An immigrant family's first Thanksgiving. A veteran's quiet afternoons with his granddaughter. A group of international students forging friendships far from home. But most of these stories will never become comics or animations, because visual storytelling has always demanded years of artistic training.

Meanwhile, comics and manga are among the most powerful storytelling media on Earth, crossing every language barrier, reaching every age group. Over a billion people read vertical comics on their phones every day. Yet almost none of them can create one.

We wanted to change that. One photo. One story. Zero artistic skill required.

What it does

Animegent is your personal AI anime studio, turning ordinary photographs into anime-style comic strips and animated short videos through natural conversation.

Upload a photo of your friends, your family, your community. Animegent transforms each person into an anime character that looks like them, writes a story around the scene, and generates a complete multi-panel manga page or even an animated video brought to life from storyboard frames. The same person keeps their consistent anime identity across every panel, every scene, and every animation.

You don't draw. You don't write prompts. You just talk to Animegent like a creative partner and watch your story come to life in real time.

How we built it

Animegent is built on Google Gemini across the entire stack. Gemini for conversational reasoning and tool orchestration, Gemini for image generation and character stylization, and Gemini as a vision-language model for scene understanding.

The core architecture is an agentic pipeline: rather than a single prompt-to-image call, the AI agent orchestrates 15+ specialized tools through Gemini function calling, including face detection, character stylization, scene analysis, script writing, comic panel generation, and animated video synthesis. Complex tasks are modeled as DAG-based plans where independent steps run concurrently and the user confirms critical decisions along the way.

The frontend streams every step live via SSE. Users see faces being detected, characters being stylized, scripts being written, and panels being rendered as it happens, making the creative process feel collaborative and magical.

What's next for Animegent

We see Animegent as a bridge between communities and their untold stories:

  • Community story campaigns: partnering with community organizations, schools, and senior centers to turn real stories into shareable manga. Imagine a grandmother who has never used social media, seeing herself as the hero of her own comic for the first time.
  • Breaking language barriers: manga is a visual language. Immigrant families, multilingual classrooms, and cross-cultural communities can share stories without needing fluent English.
  • Education: students become the characters in their own learning materials. A history lesson where you are the protagonist. A creative writing assignment where your story becomes a manga. Research shows students engage dramatically more when they see themselves represented.
  • Collaborative storytelling: multiple people contribute their photos and ideas to co-create a shared story, strengthening the bonds within any community.

We're not building a comic tool. Animegent is everyone's anime studio, giving every community, every ordinary person, the superpower to tell their own story.

Built With

Share this project:

Updates