Through Benchling's Model Hub, you can utilize your Benchling AI Credits on a variety of Scientific AI Models, both open source and proprietary.
Enable Access
To enable access to the Boltz API, ensure you have access to the Model Hub page from the sidebar. If not, contact ai@benchling.com for access. If you do not see the Boltz Lab API card on the Models page, you will need to complete the Boltz API enablement process. Email your account executive or ai@benchling.com to get started.
Add Credits to the Boltz API
Each tenant receives its own isolated Boltz Lab Workspace, which provisions its own Test and Live API keys. Credits added to your Boltz Lab Workspace are shared across your tenant, and are consumed by the API keys generated from the Boltz API Model page.
From the Boltz Lab API Model card, click "Run Model" to view the options for adding Benchling credits to your tenant's Boltz API key. Click Add to preview the number of Benchling AI credits that will be deducted and the respective amount added to your Boltz API account. Click Confirm to complete the transfer of credits.
Using the Boltz API
From the Boltz Lab API Model Card, click Generate API Keys to view your two keys, test and live. test keys only create dry runs of API calls, and do not charge for usage, whereas usage of live keys will charge based on the endpoint.
To start using the Boltz API:
- Add credits to your Boltz API account as described above. This will increase your spend limit (which starts at $0) by the amount you request.
- Explore the Boltz API Documentation and choose your method of integration (e.g. HTTP, Python SDK, etc.)
- Most endpoints contain an estimate endpoint (example) which you can use to estimate the cost of an action.
- Use the Boltz API directly with the
testorlivekey provided by Benchling. - If you exceed your spend limit, you can increase it by adding more credits from the Benchling