---
license: mit
library_name: transformers
pipeline_tag: image-text-to-text
---

This repository contains the PrismaCoder-9B model described in [PrismaCoder: Pushing the Boundaries of Open Multimodal Code Intelligence through Decomposed Rubric Synthesis](https://huggingface.co/papers/2610.20538).

Project page: https://qiushisun.github.io/PrismaCoder-Home/

Code: https://github.com/QiushiSun/PrismaCoder

PrismaCoder is a family of open multimodal code models for generating and editing charts, webpages, diagrams, documents, and animations from text or visual inputs. This checkpoint is trained on the PrismaCoder-2M dataset using supervised fine-tuning and reinforcement learning.

For more details, please refer to the [paper](https://arxiv.org/abs/2610.20538) and the [project page](https://qiushisun.github.io/PrismaCoder-Home/).