Benchmarks · Coding · Agents & computer use
LongTail-SWE: Lua
LongTail-SWE's Lua tasks: fix real GitHub issues in Lua projects. The share of attempts that pass the tests.
11 models · 11 results · data as of 2026-11-18 · longtail-swe.com ↗
Leaderboard
DeepSeek V4.1 Flash 1017 max
All results
11 rows| # | Model | Effort | LongTail-SWE: Lua | Harness |
|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash 1017 | max | 92.3% | SWE-agent-like scaffold |
| 2 | Opus 5.5 | high | 89.7% | SWE-agent-like scaffold |
| 3 | GPT-5.6 Sol | high | 89.7% | SWE-agent-like scaffold |
| 4 | GPT-6.1 Sol | high | 89.7% | SWE-agent-like scaffold |
| 5 | Kimi K3.1 | max | 84.6% | SWE-agent-like scaffold |
| 6 | GPT-5.6 Luna | xhigh | 74.4% | SWE-agent-like scaffold |
| 7 | Kimi K3 | max | 74.4% | SWE-agent-like scaffold |
| 8 | DeepSeek V4 Flash | high | 66.7% | SWE-agent-like scaffold |
| 9 | GLM 5.3 Flash | max | 59.0% | SWE-agent-like scaffold |
| 10 | Qwen 3.8 27B | xhigh | 51.3% | SWE-agent-like scaffold |
| 11 | Qwen 3.6 27B | max | 17.9% | SWE-agent-like scaffold |