Who it is for
Local operators and developers evaluating models, runtimes, accelerators, and bounded workloads.
Proposed roadmap application
Thirteenfold Compute
Local model, runtime, accelerator, workload, and evaluation management under explicit operator control.
Proposed roadmap application; name and implementation remain under review.
Why it is planned
A practical role in the family.
Example use
An operator wants to compare local models under a fixed hardware and resource ceiling before selecting a runtime profile.
Intended benefit
Could make workload behavior and resource tradeoffs easier to compare without granting models inherited tool authority.
Intended workflow
How the work would flow.
This describes the planned experience. It does not establish that these capabilities are available.
Select the approved model, runtime, workload, dataset, and resource profile.
Run the workload within declared compute and network limits.
Review metrics, logs, and evaluation evidence before any adoption decision.
Scope and relationships
Clear ownership of the work.
Data and authority boundaries
Would use explicitly selected models, datasets, workspaces, and device telemetry; it would not receive ambient filesystem access, unrestricted network access, or inherited tools.
How it fits with the family
Compute would run and evaluate models and workloads; Forge would turn source workspaces into reviewable software candidates.
Explore Thirteenfold Forge