Terminal-Bench 4.0.0 · Opus 5 / Claude Code | terminal-bench/terminal-bench | 330/330 (3 retries) | Ryan Marten | 6 | 8/26/26 |
TB 4.0.0 - GLM-5.3 / Claude Code | terminal-bench/terminal-bench | 330/330 | Ryan Marten | 8 | 8/26/26 |
Terminal-Bench 4.0.0 · Opus 5 (Max) / mini-SWE-agent | terminal-bench/terminal-bench | 330/330 (51 retries) | Ryan Marten | 13 | 8/25/26 |
Terminal-Bench 4.0.0 · GPT-5.6 Sol (Max) / Codex | terminal-bench/terminal-bench | 330/330 (19 retries) | Ryan Marten | 8 | 8/25/26 |
Terminal-Bench 4.0.0 · Opus 5 (Max) / mini-SWE-agent | terminal-bench/terminal-bench | 25/330 (10 retries) | Ryan Marten | 25 | 8/25/26 |
Terminal-Bench 4.0.0 - Oracle (10x) | terminal-bench/terminal-bench | 660/660 | Ryan Marten | 10 | 8/25/26 |
Terminal-Bench 4.0.0 · GPT-5.6 Sol (Max) / Codex | terminal-bench/terminal-bench | 19/330 | Ryan Marten | 19 | 8/25/26 |
Terminal-Bench 4.0.0 · Oracle (10x) | terminal-bench/terminal-bench | 18/660 | Ryan Marten | 18 | 8/25/26 |
Terminal-Bench 4.0.0 · Opus 5 (Max) / mini-SWE-agent | terminal-bench/terminal-bench | 18/330 | Ryan Marten | 18 | 8/25/26 |
Terminal-Bench 4.0.0 · Oracle (10x) | terminal-bench/terminal-bench | 0/660 | Ryan Marten | 0 | 8/25/26 |
Terminal-Bench 4.0.0 · GPT-5.6 Sol (Max) / Codex | terminal-bench/terminal-bench | 0/330 | Ryan Marten | 0 | 8/25/26 |
Terminal-Bench 4.0.0 · Opus 5 (Max) / mini-SWE-agent | terminal-bench/terminal-bench | 0/330 | Ryan Marten | 0 | 8/25/26 |
32500817213-cheat | vllm-deepseek-streaming | 2/2 | Ryan Marten | 1 | 8/21/26 |
32360671312-cheat | vllm-deepseek-streaming | 2/2 | Ryan Marten | 1 | 8/20/26 |
Terminal-Bench 3.0 · GLM 5.3 (Max) / Claude Code | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 15 | 8/17/26 |
Terminal-Bench 3.0 · SWE-1.7 Lightning / Devin | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 34 | 8/14/26 |
Terminal Bench 2.1 Devin CLI SWE-1.7 Lightning Full Run | terminal-bench/terminal-bench-2-1 | 445/445 | Ryan Marten | 26 | 8/13/26 |
Terminal-Bench 3.0 · Grok 4.6 / Grok Build | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 25 | 8/11/26 |
Terminal-Bench 3.0 · Opus 5 (Max) / mini-SWE-agent | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 20 | 7/22/26 |
harbor-index/judge/pr-126 | - | 397/397 | Ryan Marten | 10 | 7/20/26 |
harbor-index/judge/pr-126 | - | 397/397 | Ryan Marten | 397 | 7/20/26 |
harbor-index/judge/pr-125 | - | 399/399 | Ryan Marten | 399 | 7/20/26 |
lb-pr-124/e94e9cac-6758-4739-8645-f9d59fe0667b | harbor-index/harbor-index | 80/86 (6 retries) | Ryan Marten | 17 | 7/19/26 |
lb-pr-124/c80474fa-dcd4-44ad-bbf1-ff501b6b595a | harbor-index/harbor-index | 80/86 (6 retries) | Ryan Marten | 15 | 7/19/26 |
lb-pr-124/5e596206-ac98-4e72-b51e-c48f72927333 | harbor-index/harbor-index | 80/86 (6 retries) | Ryan Marten | 13 | 7/19/26 |
lb-pr-124/c4e230ea-4063-4466-bc25-a48288780045 | harbor-index/harbor-index | 80/87 (7 retries) | Ryan Marten | 18 | 7/19/26 |
lb-pr-123/92cbb423-7c1a-4f3b-a37b-12ae42b11aaa | harbor-index/harbor-index | 80/89 (9 retries) | Ryan Marten | 11 | 7/19/26 |
lb-pr-124/503d1ef0-6e8e-4192-b945-81f3c145f268 | harbor-index/harbor-index | 80/86 (6 retries) | Ryan Marten | 14 | 7/19/26 |
harbor-index/judge/pr-72 | - | 400/400 | Ryan Marten | 400 | 7/17/26 |
harbor-index/judge/pr-72 | - | 114/400 | Ryan Marten | 114 | 7/17/26 |
harbor-index/judge/pr-72 | - | 378/400 | Ryan Marten | 378 | 7/17/26 |
Terminal-Bench 3.0 · GLM 5.2 (Max) / Claude Code | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 134 | 7/17/26 |
harbor-index/judge/pr-72 | - | 400/400 | Ryan Marten | 400 | 7/17/26 |
harbor-index/judge/pr-72 | - | 400/400 | Ryan Marten | 293 | 7/17/26 |
harbor-index/judge/pr-72 | - | 10/10 | Ryan Marten | 0 | 7/14/26 |
Terminal-Bench 3.0 · Grok 4.5 (xHigh) / Cursor CLI | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 86 | 7/12/26 |
Terminal-Bench 3.0 · GPT-5.6 Luna (Max) / Codex | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 27 | 7/12/26 |
Terminal-Bench 3.0 · Sonnet 5 (Max) / Claude Code | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 65 | 7/12/26 |
Terminal-Bench 3.0 · GPT-5.6 Terra (Max) / Codex | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 11 | 7/12/26 |
Terminal-Bench 3.0 · Opus 4.8 (Max) / Claude Code | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 28 | 7/12/26 |
Terminal-Bench 3.0 · Fable 5 (Max) / Claude Code | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 36 | 7/12/26 |
Terminal-Bench 3.0 · GPT-5.6 Sol (Max) / Codex | terminal-bench/terminal-bench | 370/370 | Ryan Marten | 23 | 7/12/26 |
tb-2-1/judge/pr-106 | - | 363/363 (1 retry) | Ryan Marten | 0 | 7/11/26 |
tb-2-1/judge/pr-69 | - | 294/294 (2 retries) | Ryan Marten | 0 | 7/8/26 |
tb-2-1/judge/pr-67 | - | 261/261 | Ryan Marten | 0 | 7/8/26 |
tb-2-1/judge/pr-47 | - | 445/445 | Ryan Marten | 54 | 7/8/26 |
tb-2-1/judge/pr-45 | - | 444/444 | Ryan Marten | 51 | 7/8/26 |
2026-07-08__06-28-00 | harbor-index/harbor-index-1.0 | 404/410 (85 retries) | Ryan Marten | 66 | 7/8/26 |
harbor-index/judge/pr-8 | - | 394/394 | Ryan Marten | 8 | 7/8/26 |
2026-07-07__20-10-30 | harbor-index/harbor-index-1.0 | 356/410 (99 retries) | Ryan Marten | 83 | 7/7/26 |