AI EXECUTIONS

Reusable execution capabilities for developer applications.

AI Executions let applications discover and invoke reusable AI workload capabilities through Tenzarch while relying on the same execution network used by other workload sources.

DEVELOPER WORKLOADS

AI Executions create trackable network workloads.

An application selects an available AI Execution, supplies the required input, and submits it through the developer interface. Tenzarch creates execution state and returns references that the application can track asynchronously.

Discovery

Find execution capabilities exposed by the current platform.

Input

Provide the capability-specific workload input.

Submission

Create a Job rather than treating execution as an opaque synchronous call.

Tracking

Follow lifecycle, logs, telemetry, and final outcome.

CURRENT HTTP FLOW

The current backend contract preserves its existing route names.

Discover
GET /api/marketplace/services
x-api-key: dev_...
Execute
POST /api/marketplace/services/:id/execute
x-api-key: dev_...
Idempotency-Key: <unique-key>
Content-Type: application/json

{
  "input": {
    "prompt": "Analyze this dataset and return concise findings.",
    "format": "summary"
  }
}