Skip to main content

Documentation Index

Fetch the complete documentation index at: https://hexelstudio.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

Instances are running deployments of registered agents. Each instance receives a permanent endpoint that does not change across redeployments.

​
Deploy

from hexel import Hexel

client = Hexel(api_key="your_api_key")

instance = client.compute.instance.deploy(
    "agent-id",
    env={"MODEL": "gpt-4", "API_KEY": "sk-xxx"}
)

print(instance["endpoint"])

​
Permanent endpoints

Each unique configuration produces a unique, permanent URL: 
Same agent + same environment → same URL
Same agent + different environment → different URL
The URL is derived from a hash of the agent identity and configuration. Redeploying with identical settings returns the same endpoint.

​
Call your agent

Once deployed, call your agent directly at its endpoint: 
# Streaming response
curl -N https://zeal-isle-a620.compute.hexelstudio.com/stream \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query": "Hello"}'

# Blocking response
curl https://zeal-isle-a620.compute.hexelstudio.com/invoke \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query": "Hello"}'

​
Manage instances

instances = client.compute.instance.list()

client.compute.instance.stop("deployment-id")

client.compute.instance.redeploy("deployment-id")

client.compute.instance.delete("deployment-id")

​
States

StateDescription
DeployingThe instance is starting
RunningHealthy and serving traffic
StoppedManually stopped by the user
FailedThe instance could not start
RestartingRecovering automatically after a failure
Instances recover automatically if the underlying infrastructure fails. The endpoint remains the same throughout the lifecycle.