CineAI

A multi-modal AI director's chair for seamless storytelling. Transforming ideas into cinematic sequences with generative AI.

Comment

Inspiration : We stumbled upon this idea during brainstorming session

What it does : CINE AI is an autonomous, multimodal studio assistant that intelligently routes your prompts to generate screenplays, concept art, and videos.

How we built it : We built a Streamlit frontend integrated with the official Google GenAI SDK, using Gemini 2.5 Flash as an autonomous router to orchestrate text, image, and Veo video generation.

Challenges we ran into: We tackled strict Pylance type-checking errors, built an LLM translator to stop image models from failing on complex screenplays, and migrated entirely to the official Google GenAI SDK.

Accomplishments that we're proud of : Despite being a beginner, we built a fully autonomous AI studio, got familiar with Google's SDK, and integrated complex Veo video generation.

What we learned : We learned to build a fully autonomous AI agent, Google GenAI SDK, engineering dynamic prompts and handling complex multimodal Streamlit states.

What's next for CineAI: Integrating AI audio generation for sound design, building a timeline editor

Built With

  • css
  • google-cloud-platform-(gcp)
  • google-genai-sdk
  • html
  • python
  • python-dotenv
  • streamlit
  • vertex-ai
Share this project:

Updates