Agent deployments on Beamlit have an inference endpoint which can be used by external consumers to request an inference execution. Inference requests are then routed on the Global inference Network based on the deployment policies associated with your agent deployment.

​
Inference endpoints

Whenever you deploy an agent on Beamlit, an inference endpoint is generated on Global inference Network.

The inference URL looks like this:


run.beamlit.dev/{your-workspace}/agents/{your-agent}

There is one distinct endpoint for each agent deployment, i.e. for each combination of a agent and an environment on which it is deployed.

For example, if you have one version of agent “your-agent” deployed on the production environment and one version deployed on the development environment:

  • run.beamlit.dev/{your-workspace}/agents/{your-agent}?environment=production will call the production deployment
  • run.beamlit.dev/{your-workspace}/agents/{your-agent}?environment=development will call the development deployment

If you do not specify the environment in the inference request, it will call the production environment by default. If the agent is not deployed on the production environment, it will return an error.

​
Endpoint authentication

By default, agents deployed on Beamlit aren’t public. It is necessary to authenticate all inference requests, via a bearer token.

The evaluation of authentication/authorization for inference requests is managed by the Global inference Network based on the access given in your workspace.

Making an agent publicly available is not yet available. Please contact us at support@beamlit.com if this is something that you need today.

​
Make an inference request

​
Beamlit API

Make a POST request to the inference endpoint for the agent deployment you are requesting, making sure to fill in the authentication token:

curl 'https://run.beamlit.dev/YOUR-WORKSPACE/agents/YOUR-agent?environment=YOUR-ENVIRONMENT' \
  -H 'accept: application/json, text/plain, */*' \
  -H 'x-beamlit-authorization: Bearer YOUR-TOKEN' \
  -H 'x-beamlit-workspace: YOUR-WORKSPACE' \
  --data-raw $'{"inputs":"Enter your input here."}'

Read about the API parameters in the reference.

​
Beamlit console

Inference requests can be made from the Beamlit console from the agent deployment’s workbench page.

OverviewOverview