SWE-2 Flash: benchmark results
Provider: Other. Released 2026-11-04. Access: Interface.
Aggregate Rating 1811 ± 36, rank #69 of 10800 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI Coding Daily (Devin) - Python Code Quality | 17.4 | Python Code Quality (max 20) points, LLM-judged rubric score | 100 |
| Cognition SWE-2 Flash Release - FrontierCode 2.0 | 24 | Score (%) | 100 |
| LLM Stats (Terminal-Bench 2.1) | 89.2 | Score (%) | 94.3 |
| LLM Stats Score | 43.78 | LLM Stats Score (conservative rating) | 86.6 |
| LLM Stats (SWE Atlas - Codebase QnA) | 50 | Score (%) | 57.1 |
| LLM Stats (FrontierCode 1.1) | 45.5 | Score (%) | 56.5 |
| AI Coding Daily (Devin) - Laravel Code Quality | 17.4 | Laravel Code Quality (max 20) points, LLM-judged rubric scor | 50 |
| LLM Stats (DeepSWE 1.1) | 60.1 | Score (%) | 26.5 |
| LLM Stats (ExploitBench) | 34.8 | Score (%) | 22.2 |
| AI Coding Daily (Devin) - React-TS Code Quality | 17.33 | React-TS Code Quality (max 20) points, LLM-judged rubric sco | 0 |
| AI Coding Daily (Devin) - Total | 68.55 | Total points (max 90) | 0 |
| Cognition SWE-2 Flash Release - DeepSWE 1.1 | 60.1 | Score (%) | 0 |
Interactive version: theaggregate.ai/model?slug=swe-2-flash · How It Works · Data refreshed daily, snapshot 2026-11-13.