New

  • Red Team: Red Team is now available online, with tools to create and run assessments, review results, and generate reports. The open-source Red Team engine is also available on PyPI.
  • Experiments: added filters for datasets, prompts, evaluators, and models, along with model, prompt, and evaluator details directly in experiment lists.
  • Dataset automation: automatic insert workflows now support trace datasets.
  • Dashboard autosave: named dashboard views now save automatically, making it easier to preserve dashboard changes as you work.
Red Team campaign report showing an overall grade of F, counts for confirmed findings, resistance rate, tested rounds, and duration, and a findings list with critical severity labels.
Red Team campaign report

Improved

  • Experiments: added starring, provider icons for models, and server-side table sorting for more useful experiment management at scale.
  • Dashboards: improved synced chart-hover markers and comparison labels so tooltips and date ticks consistently respect the selected timezone.
  • Home dashboard: metrics refresh automatically in the background and update more reliably when returning to the page or changing the time range.
  • Metrics: charts now support copy-to-clipboard.
  • Behaviors: improved behavior detection accuracy and reduced false positives.

Fixed

  • Traces: span-level filters are now preserved when viewing grouped traces and threads, so results remain accurate.
  • Security: strengthened protections.