SkufosaurRU
Menu
← All news

Analysis

80%

Terminal-Bench 2.1 repaired 28 of 89 tasks. Compare agents on the same version.

Before comparing agents, check the benchmark version

Terminal-Bench 2.1 repaired 28 of 89 tasks. Compare agents on the same version.

What to do

  1. Record the dataset, harness, agent and model versions.
  2. Run shortlisted agents against the same tasks and budgets.
  3. Compare results alongside cost and failed tasks.

Why the version matters

The May 6 release repaired 28 of 89 tasks. In the comparison published by the benchmark authors, Claude Code with Opus 4.6 gained 12.1 percentage points. Several other pairs gained less or declined slightly.

A leaderboard without a version is a stopwatch without a start button.

Primary source and results by configuration

Plus if it helped