Queued or running
{% extends "layout.html" %}
{% block title %}Evaluations · cua-speedrun{% endblock %}
{% block content %}
Run status, results, and execution history. Queued or running Results available All runs in this accountEvaluations
Add a token pair for remote Modal runs. Fully local evaluations do not require one.
Remote placements run in your Modal workspace. Local placements use the evaluator host. Token secrets are encrypted at rest and are not displayed after saving.
{% if modal_token_id %}Token ID: {{ modal_token_id[:6] }}…{{ modal_token_id[-4:] }}
Save a variable once, then select it when starting an evaluation. Values are encrypted, hidden after saving, and available only to that evaluation's init.py and agent.py. Saving an existing name replaces its value.
{{ name }}Start with a template or upload a submission.
New evaluationSign in to run and view evaluations.
Sign in