Benchmark
Ramp SWE-Bench v2 (Bash)
Coding benchmarks compare models on programming tasks. Scores are percentages.
All coding benchmarksScores from LLM Stats data.
- Models measured
- 7
- Scores recorded
- 7
- Sources
- 1
- IQ weight
- 0.58
Score distribution
- 56.8%
- Mean
- 54.6%
- Median
- 11.6%
- Std dev
- 45.4%
- 10th percentile
- 72.4%
- 90th percentile
- 42.6%
- Min
This benchmark has a weight of 0.58 for chat models, measured over 7 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT-6.2 Astra· LLM Stats data | 75% |
| 2 | GPT-6.2 Sol· LLM Stats data | 72.4% |
| 3 | Grok 4.7· LLM Stats data | 55.4% |
| 4 | GPT-6 Astra· LLM Stats data | 54.6% |
| 5 | GPT-6.1 Sol· LLM Stats data | 52% |
| 6 | Kimi K3.1 Pro· LLM Stats data | 45.4% |
| 7 | GPT-6 Luna· LLM Stats data | 42.6% |
Data credits: Data by LLM Stats