---
id: "b921"
language: "en"
library: "tools"
official_url: "https://github.com/sokrypton/ColabFold"
profile_updated_on: "2026-10-22"
---

# ColabFold

Combines MMseqs2 sequence searches with protein-structure models through Colab notebooks and local batch workflows.

- Company / organization: ColabFold project
- Category: Model development
- Official website: [https://github.com/sokrypton/ColabFold](https://github.com/sokrypton/ColabFold)
- Profile updated: 2026-10-22


## Overview

ColabFold packages sequence alignment and structure prediction into reusable workflows, with AlphaFold2, complex-prediction and ESMFold notebooks plus local batch instructions. It runs predictions rather than only searching stored results. Colab depends on available GPUs; the public MSA service requests serial queries from one IP. Local use can still involve external searches. MIT code licensing does not automatically cover every model, weight or database.

## Official sources and references

- [Official website](https://github.com/sokrypton/ColabFold)
- [Documentation](https://github.com/sokrypton/ColabFold/blob/main/README.md)

## Reading & subscriptions

- [Interactive profile](../../../products/colabfold--b921.html)
- [Read](../../../read/en/products/colabfold--b921.html)
- [JSON](../../../read/en/products/colabfold--b921.json)

Descriptions summarize public information; consult official websites for features, pricing and regional availability.