agents benchmark
DeepSWE 1.1
DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.
Updated Aug 11, 2026
Models20
Model coverage20
MetricScore
EvidenceB
DeepSWE 1.1 Ranking
Higher score ranks better on this benchmark.
| 01 | OP | 73.0% | 100.0% | 20 | B | |
| 02 | AN | 70.0% | 94.7% | 20 | B | |
| 03 | OP | 70.0% | 89.5% | 20 | B | |
| 04 | MA | 69.0% | 84.2% | 20 | B | |
| 05 | AN | 68.8% | 79.0% | 20 | C | |
| 06 | OP | 67.0% | 73.7% | 20 | B | |
| 07 | OP | 67.0% | 68.4% | 20 | B | |
| 08 | ME | 59.3% | 63.2% | 20 | C | |
| 09 | AN | 59.0% | 57.9% | 20 | B | |
| 10 | AC | 56.6% | 52.6% | 20 | C | |
| 11 | AN | 54.0% | 47.4% | 20 | B | |
| 12 | XA | 54.0% | 42.1% | 20 | B | |
| 13 | ME | 53.0% | 36.8% | 20 | B | |
| 14 | OP | 52.0% | 31.6% | 20 | B | |
| 15 | GO | 49.0% | 26.3% | 20 | B | |
| 16 | ZA | 44.0% | 21.1% | 20 | B | |
| 17 | GO | 37.0% | 15.8% | 20 | B | |
| 18 | MA | 31.0% | 10.5% | 20 | B | |
| 19 | AN | 30.0% | 5.3% | 20 | B | |
| 20 | GO | 12.0% | 0.0% | 20 | B |
DeepSWE 1.1 Score Distribution
A closer view of the leading scores on this benchmark.
DeepSWE 1.1 Highlights
The leading models and scores on this benchmark.
What is DeepSWE 1.1?
What DeepSWE 1.1 measures and how its scores work.
DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- DeepSWE 1.1
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- deepswe-1.1|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about DeepSWE 1.1.
Which model scores highest on DeepSWE 1.1?
GPT-5.6 Sol is currently ranked first with 73.0%.
What does DeepSWE 1.1 measure?
DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
20 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.