| Failure Recovery via Belief Updates and World Modeling for Long-Horizon Agents | | 66.6 | BeliefWeaver (Terminus-2) | 2026-10-13 |
| Terminal-Bench Leaderboard | ✓ Link | 42.7 | Claude Opus 5 (Max) (mini-SWE-agent) | 2026-07-24 |
| Terminal-Bench Leaderboard | ✓ Link | 34.59 | GPT-5.6 Sol (Max) (Codex) | 2026-07-09 |
| Terminal-Bench Leaderboard | ✓ Link | 34.05 | Claude Fable 5 (Max) (Claude Code) | 2026-06-09 |
| DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression | ✓ Link | 30.0 | DeepSeek-V4.1-Flash Max (DeepSeek Harness (Minimal)) | 2026-09-17 |
| GLM-5.3: Frontier Coding with Emergent Cyber Capabilities | ✓ Link | 28.3 | GLM-5.3 (Claude Code) | 2026-08-14 |
| Terminal-Bench Leaderboard | | 26.49 | Grok 4.6 (High) (Grok Build) | 2026-08-12 |
| Terminal-Bench Leaderboard | | 21.08 | Claude Opus 4.8 (Max) (Claude Code) | 2026-05-28 |
| Terminal-Bench Leaderboard | ✓ Link | 20.81 | GPT-5.6 Terra (Max) (Codex) | 2026-07-09 |
| Terminal-RSI: Recursive Self-Improving Terminal Agents via Recursive Self-Rewrite and Policy-Refine | | 16.95 | Terminal-RSI-27B | 2026-10-15 |
| Terminal-Bench Leaderboard | | 15.68 | Grok 4.5 (XHigh) (Cursor CLI) | 2026-07-08 |
| Introducing Gemini 3.7 Flash | ✓ Link | 14.9 | Gemini 3.7 Flash (MiniSWE-Agent) | 2026-08-13 |
| Terminal-Bench Leaderboard | | 14.59 | Claude Sonnet 5 (Max) (Claude Code) | 2026-06-30 |
| Terminal-Bench Leaderboard | ✓ Link | 14.32 | GPT-5.6 Luna (Max) (Codex) | 2026-07-09 |
| Terminal-Bench Leaderboard | ✓ Link | 4.59 | GLM-5.2 (Max) (Claude Code) | 2026-06-16 |