HelixFold3
Open-source AlphaFold3-class biomolecular structure prediction for proteins, nucleic acids and ligands.
About
HelixFold3 is the PaddleHelix team's open attempt to replicate AlphaFold3's capabilities: it predicts the structures of proteins, protein complexes, nucleic acids and small-molecule ligands from sequence and SMILES inputs. The code and model parameters were released in August 2024 for non-commercial academic research, and the team runs a public web server plus a paid SDK API for high-throughput academic and commercial use. A follow-up, HelixFold3.2, shipped in July 2025 with reported gains on protein-related tasks.
In the authors' own reporting, HelixFold3 matches AlphaFold3-level accuracy on conventional ligands, nucleic acids and proteins, with over 90% success on ligand predictions in the PoseBusters benchmark. Because the weights are genuinely downloadable, it has become one of the few practical routes for labs that want AlphaFold3-style complex prediction they can run and modify themselves — the trade-off being the non-commercial license on the initial release.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
- Paper-reported (arXiv 2408.16975): over 90% success on ligand predictions on PoseBusters; competitive scores on RNA and DNA structure evaluation
- No independent third-party benchmarks published
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (TIER metrics, metrics-v1). Evidence: technical report on arXiv (2408.16975) with PoseBusters/RNA/DNA evals; open code + parameters on GitHub (PaddlePaddle/PaddleHelix); maintained web server and SDK, with HelixFold3.2 released July 2025. Strengths: genuinely downloadable AlphaFold3-class weights, active maintenance, real academic uptake. Docked for: (1) benchmark numbers are author-reported with no independent replication found; (2) initial release is non-commercial only — commercial users must use the paid API. License: non-commercial (academic research) for the initial open release.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 6.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Technical report arXiv 2408.16975; >90% ligand success on PoseBusters per authors
- · Code + parameters released Aug 2024 (PaddlePaddle/PaddleHelix); HelixFold3.2 released Jul 2025
- · Public web server + paid SDK API maintained