Loading…
Bio platform
Benchling Model Hub
Managed structure prediction on registry data inside the electronic lab notebook, and on uploaded sequences from a standalone benchling.ai account.
- Models served
- 14
- Offers
- Inference
- Coverage
- Protein 13Small molecule 5DNA & Gene 1RNA 1
- Deployment
- In-app or standalone · cloud or VPC
- Pricing
- Free tier · subscription + credits
Models you can run on Benchling Model Hub
Most cited first38.6K—14.8K
Protein structure prediction model that folds amino acid sequences into 3D structures with atomic accuracy, scoring a median GDT of 92.4 at CASP14.
Protein
571—4.2K
Open model that jointly predicts biomolecular structure and small-molecule binding affinity, approaching FEP+ accuracy in seconds on a single GPU.
Protein
466—4.2K
Open-source structure prediction model for proteins, nucleic acids, and small molecules, trained on public data to AlphaFold3-level accuracy.
Protein
461—3.4K
Trainable, open-source reimplementation of AlphaFold2 for protein structure prediction that matches its accuracy and runs 3-5x faster.
Protein
442—2K
Biomolecular structure prediction foundation model covering proteins, small molecules, DNA, RNA, and glycans in a single diffusion framework.
Protein
How access works
Benchling.ai is self-serve: an account is created directly, with no Benchling subscription, sales conversation or credit card, and predictions run on uploaded sequences under a free usage allowance. Inside the platform, Model Hub is enabled per tenant, runs on registered entities, and is absent from Validated Cloud. Identity there comes from the customer's own provider over SSO or OIDC, runs are written to the tenant audit trail, and inference stays on managed compute or in a customer VPC.
- Self-serve
- Free tier
- Academic pricing
- SSO / SAML
- SOC 2 Type II
- ISO 27001
- GDPR DPA
- Region selectable
What it costs
Two billing shapes. On benchling.ai the service is provided at no charge, with no card required and a published allowance of up to $1,000 of free usage a year, based on typical usage patterns. Inside the Benchling platform, model and agent runs draw on a credit system, and credits are included with every subscription; consumption per run varies by model. No rate list is published on either route, and a tenant that exhausts its credits waits until they refresh on the first of the month.
Last verified
