Public KG record
DeepSWE v1.1
Recorded overview
Benchmark: Astra achieves 74%.
An evidence-backed entity from the project knowledge graph; detailed tool metadata is still being added.
Recorded details
- Benchmarks: benchmark for software engineering
- Benchmarks: DeepSeek-V4.1-Flash 在 DeepSWE v1.1 上超过 Opus-5 与 GPT-5.6 Sol
- Benchmarks: Sol 在 max effort 下得分 68.8%
- Benchmarks: Luna 在 max effort 下得分 66.6%
- Benchmarks: 用于编码评测
- Benchmarks: 评测长周期软件工程
- Benchmarks: 用于评估 Astra，并比较启用与未启用水印的输出
- Benchmarks: 得分73.7
Evidence coverage
The KG currently records 6 known information dimensions and has 11 undocumented dimensions, linked to 18 public source articles. These counts describe record coverage, not model performance or verified fit.
Related public records
- Astra — evaluated on
- DeepSeek-V4.1-Flash-1017 — evaluated on
- DeepSeek-V4.1-Flash — evaluated on
- Gemini 3.7 Flash — evaluated on
- Gemini 3.8 Flash — evaluated on