recAMPLIFY
recAMPLIFY is a protein structure & design tool whose code is published by chandar-lab on GitHub, first public in 2026, described in arXiv (2026). Chandar Lab's recursive masked protein language model: a weight-tied recurrent core refines residue embeddings over several cycles, so you can spend more compute per sequence at inference without adding parameters.
| Category | Protein structure & design |
|---|---|
| Developer (GitHub) | chandar-lab |
| Task | Masked protein language modeling and residue embeddings |
| Interface | Transformers (trust_remote_code) |
| Website | huggingface.co/chandar-lab/recamplify |
| GitHub | chandar-lab/AMPLIFY |
| Released | 2026-10-19 (source) |
| Paper | https://arxiv.org/abs/2610.26747 · arXiv (2026) |
| Code license | MIT |
| License note | MIT (Hugging Face model card; AMPLIFY GitHub repo) |
| What this allows | Permissive (commercial OK) Use, modify and ship it inside closed-source commercial products; keep the license text and copyright notices. What do these mean? |
| License source | huggingface.co/chandar-lab/recamplify · checked 2026-10-26 |
| Pricing | Open source |
| Tags | #protein-language-model#embeddings#recursive-transformer |
| Last checked | · links verified |
Can I use recAMPLIFY commercially?
Yes. The recorded license is permissive, so closed-source commercial use is allowed if you keep the notices.
- Code: MIT
- Source: huggingface.co/chandar-lab/recamplify · license checked 2026-10-26
What the license types mean · not legal advice.
Specs
| Task | Masked protein language modeling and residue embeddings |
|---|---|
| Modalities | — |
| MSA required | — |
| Open weights | Yes (Hugging Face chandar-lab/recamplify, 11 checkpoints) |
| GPU need | — |
| GitHub stars | 136 |
| Interface | Transformers (trust_remote_code) |
Compare with…
Pick 1–3 other Protein structure & design tools to compare side by side with recAMPLIFY.
Related
DeepMind's open-source protein structure prediction model (monomer and multimer).
Structure prediction of proteins, nucleic acids, ligands and their complexes; hosted web server plus open code, weights on request for non-commercial use.
Open biomolecular interaction models (Boltz-1, Boltz-2) for complex structure and binding-affinity prediction.