Hugging Face Releases DNA-Chat AI Model Carbon-X
Original: Hugging Face’s New AI Biology Model Lets You Chat With DNA
Why This Matters
Conversational genomics tools could widen access to DNA analysis well beyond specialist labs.
Hugging Face launched Carbon-X on Oct 22, an open-source model trained on 300B tokens of DNA sequences and natural language, letting researchers query genetic mutation effects via conversational prompts on a standard laptop.
Hugging Face released Carbon-X, an open-source biology model built with researchers at Zhongguancun Academy in China and the University of Naples Federico II in Italy. Unlike prior tools such as Google DeepMind's AlphaGenome—which outputs numeric predictions and demands heavy compute—Carbon-X accepts natural-language questions about DNA and returns plain-English answers, running locally on a laptop.
The model began as an existing language model whose vocabulary was extended to read DNA in six-letter chunks. Training ran across 300 billion tokens combining genomic sequences from mammals, vertebrates, invertebrates, plants, fungi, and protozoa, alongside English text and descriptions linking DNA to biological function. Viruses were deliberately excluded to limit misuse potential.
Users can query segments of roughly 50,000 base pairs at a time—not full genomes. Senior research scientist Lewis Tunstall framed the target use cases as drug design, genetic therapy development, and potentially helping patients interpret their own genomic data. "We've unlocked a new capability in language models, which is fundamentally about being able to chat with DNA," he said.
The release comes weeks after Nvidia agreed to acquire Hugging Face for nearly $13 billion. Carbon-X joins a crowded but still-maturing field that includes Arc Institute and Nvidia's Evo 2, which can generate whole genomes but lacks Carbon-X's conversational interface.