# TranscriptFormer and rBio (CZI virtual cell models)

Source: https://onco.cc/technologies/transcriptformer/  
OnCo record `transcriptformer` (Technology). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

CZI's open cross-species cell models and a reasoning model trained on them.

## Summary

TranscriptFormer is the Chan Zuckerberg Initiative's open generative transformer over single cells, trained on 112M cells across 12 species so it can compare cell states across organisms. Its companion rBio uses the outputs of virtual cell models as a teacher to train a language model that reasons about gene perturbations in plain language, a distillation of a simulator into a conversational model. Both are released through the CZI Virtual Cells Platform and are intended for biologists who want open, cross-species tools rather than proprietary ones. The work is early, published in 2025, and independent evaluation of how well rBio's reasoning tracks experimental results is still to come. For a newcomer: these are free models from CZI that map cells across species and let you ask a chatbot what a gene knockout might do.

## Fields

- Kind: Technology
- Status: emerging
- Last checked: 2026-09-08
- Tags: foundation-model; virtual-cell
- Principle: TranscriptFormer is a generative transformer over cells, distilled into an LLM.
- Since: 2025
- Strengths: Open, cross-species
- Limitations: Early

## Sources

- CZI Virtual Cells Platform: https://virtualcellmodels.cziscience.com

## Connected records

- fronts: [AI & Computation](https://onco.cc/fronts/ai-computation/)
- companies: [Chan Zuckerberg Initiative (Biohub)](https://onco.cc/companies/chan-zuckerberg-initiative/)
- roadmaps: [Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell](https://onco.cc/roadmaps/virtual-cell/)

---
JSON: https://onco.cc/api/v1/entities/transcriptformer.json