| HLE(vendor card, max effort, no tools) | 64.4% |
| Terminal-Bench 4.0(vendor-reported, xhigh effort; Artificial Analysis measures 59.6 in its own harness) | 66.4% |
| HLE (w/tools)(vendor-reported, with tools) | 67.7% |
| Chartography (w/tools)(vendor-reported, with tools) | 89% |