Open-source DNA foundation model processing nearly 400,000 base pairs, able to continue sequences, score variants and predict protein structure. Claimed 275x faster than Evo 2 medium; sizes from 500M to 8B parameters.
Carbon
by Hugging Face
Open-source DNA foundation model family from Hugging Face, 500M to 8B parameters, with ~400k base-pair context.
Updated EU status checked Last shown
Carbon at a glance
- What is Carbon?
- Carbon is a model from Hugging Face. Open-source DNA foundation model family from Hugging Face, 500M to 8B parameters, with ~400k base-pair context.
- Who makes Carbon?
- Carbon is made by Hugging Face.
- Does Carbon work in the EU?
- Unverified. No explicit evidence found. Hosted on Hugging Face and described as open-source, but license terms were not verified. Last verified 14 Nov 2026.
- How much does Carbon cost?
- Open source. Described as open-source; free demo on Hugging Face Spaces. Exact license not confirmed.
- Where was Carbon shown?
- In 1 creator video by AI Search. Most recently by AI Search on 24 May 2026, at 10:29 in “AI scientist, DNA editors, AI NPCs, new Qwen, open-source robots, new video editors: AI NEWS”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Carbon is a DNA foundation model released by Hugging Face (HuggingFaceBio). According to a creator summary, it processes sequences of nearly 400,000 base pairs and is offered in sizes from 500M to 8B parameters.
Reported capabilities include continuing DNA sequences, scoring genetic variants and predicting protein structure. The same source claims it runs about 275x faster than Evo 2 medium; this figure is unverified here and no official benchmarks were available.
A demo is hosted as a Hugging Face Space. No official documentation, paper or license text could be fetched, so license terms and release details are not confirmed.
- Context of nearly 400,000 base pairs
- Model sizes from 500M to 8B parameters
- Sequence continuation (generation)
- Variant effect scoring
- Protein structure prediction
- Claimed 275x faster than Evo 2 medium (unverified)
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Similar AI tools
Closest neighbours by meaning, not just by category.
Carbon-X
Model · Hugging Face
Open 14.8B language model from Hugging Face that reads raw DNA in chat and answers in English, built on Qwen3-14B.
MAMMAL
Model
Multimodal biomedical foundation model for drug discovery across small molecules, proteins, antibodies and gene expression.
LOGOS
Model · Alibaba Tongyi Lab
Autoregressive scientific model that tokenizes proteins, molecules, materials, antibodies and reactions into one shared grammar.
Botanic 1
Model · Google DeepMind
Plant genomic language model from Living Models, trained on 320 plant species and paired with Gemma 4 for crop genetics.
Kromoss
Model
Community full fine-tune of MOSS-Image-Base at 1024 px, with optional 8-step turbo LoRAs; open weights under Apache-2.0.
Cosmos 3
Model · NVIDIA
NVIDIA's open omni-model family for physical AI: world simulation, reasoning and action generation for robots and AVs.