Public KG record
GPT-6.1 Sol
Recorded overview
Powers AI agents, balancing reasoning, speed, and cost.
An evidence-backed entity from the project knowledge graph; detailed tool metadata is still being added.
Recorded details
- Benchmarks: DeepSWE v1.1表现与Astra相当
- Benchmarks: OSWorld 2.0离线集得分距Astra 2.1点
- Benchmarks: Terminal-Bench Science 0.1每任务成本$5.47
- Benchmarks: 144 次研究中的浏览器工作流测试
- Benchmarks: 在500项评分判断中，与资深人工审核者的一致率为88.6%
- Benchmarks: pandoc pass rate 从 1 agent 的 56.19% 升至 32 agents 的 79.59%
- Benchmarks: Agents’ Last Exam v1：58.6%
- Benchmarks: MMMU-Pro (Vision)：86.3%
Evidence coverage
The KG currently records 15 known information dimensions and has 4 undocumented dimensions, linked to 25 public source articles. These counts describe record coverage, not model performance or verified fit.
Related public records
- Agensh — uses
- ChatGPT Work — supports
- ChatGPT Work — uses
- CodeRabbit security agent — uses
- Droid 1.0 — supports