82.2%Checked A person checked this number against its source on 7 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
93.0%Checked A person checked this number against its source on 7 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
90.2%Checked A person checked this number against its source on 7 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Terminal-Bench 4.0launch post names both Artificial Analysis and Vals as sources of its Terminal-Bench v4.0 scores
13.1%Checked A person checked this number against its source on 7 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
45.6%Checked A person checked this number against its source on 8 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
DeepSWE v1.1Run by Datacurve · effort 0.99; 1M-token trajectory limit
72.1%Corroborated The same number is printed in 1 other lab document: 1
85.1%Checked A person checked this number against its source on 8 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Run by Artificial Analysis · Terminus harness
81.3%Checked A person checked this number against its source on 7 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
MCP AtlasPublic set (500 tasks) · Run by Scale AI · public set; effort 0.99; 1M-token trajectory limit
85.0%Corroborated The same number is printed in 1 other lab document: 1
MCP AtlasFull set (1,000 tasks) · Run by Scale AI · full set; effort 0.99; 1M-token trajectory limit
85.4%Corroborated The same number is printed in 1 other lab document: 1
