Terminal Agentic Trajectory Generation on TerminalBench 1.0
56.25ScoreDeepSeek-V3.2
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-V3.2Model Size=685B, Agent Scaffolds=Terminus-22026.02 | 56.25 | |
| GPT-5.2Agent Scaffolds=Terminus-22026.02 | 54.38 | |
| Gemini-3-Flash-PreviewAgent Scaffolds=Terminus-22026.02 | 53.72 | |
| Claude-Sonnet-4.5Agent Scaffolds=Terminus-22026.02 | 51 | |
| GLM-4.7Model Size=358B, Agent Scaffolds=Terminus-22026.02 | 48.75 | |
| Claude-Opus-4.5Agent Scaffolds=Terminus-22026.02 | 47.5 | |
| Gemini-3-Pro-PreviewAgent Scaffolds=Terminus-22026.02 | 46.35 | |
| Kimi-K2-InstructModel Size=1T, Agent Scaffolds=Terminus-22026.02 | 44.59 | |
| Minimax-M2.1Model Size=229B, Agent Scaffolds=Terminus-22026.02 | 42.5 | |
| Qwen3-Coder-480B-A35B-InstructModel Size=480B, Agent Scaffolds=Terminus-22026.02 | 37.5 | |
| GPT-5.1Agent Scaffolds=Terminus-22026.02 | 35.5 | |
| TerminalTraj-32BModel Size=32B, Agent Scaffolds=Terminus-22026.02 | 35.3 | |
| DeepSeek-V3.1-Nex-N1Model Size=685B, Agent Scaffolds=OpenHands2026.02 | 31.56 | |
| TerminalTraj-14BModel Size=14B, Agent Scaffolds=Terminus-22026.02 | 28.91 | |
| Qwen3-32B-Nex-N1Model Size=32B, Agent Scaffolds=OpenHands2026.02 | 28.75 | |
| Qwen3-30B-A3B-Nex-N1Model Size=30B, Agent Scaffolds=OpenHands2026.02 | 25 | |
| Qwen3-Coder-30B-A3B-InstructModel Size=30B, Agent Scaffolds=Terminus-22026.02 | 23.8 | |
| TerminalTraj-7BModel Size=7B, Agent Scaffolds=Terminus-22026.02 | 23.01 | |
| Qwen3-235B-A22B-InstructModel Size=235B, Agent Scaffolds=Terminus-22026.02 | 15 | |
| Qwen3-32BModel Size=32B, Agent Scaffolds=OpenHands2026.02 | 11.25 | |
| Qwen2.5-Coder-7B-InstructModel Size=7B, Agent Scaffolds=Terminus-22026.02 | 6.25 | |
| Qwen2.5-Coder-14B-InstructModel Size=14B, Agent Scaffolds=Terminus-22026.02 | 6.25 | |
| Qwen2.5-Coder-32B-InstructModel Size=32B, Agent Scaffolds=Terminus-22026.02 | 5 |