Office Application Question Answering on OfficeQA (held-out test)
72.1Score (%)SkillOpt
Evaluation results
| Method | Links | |
|---|---|---|
| SkillOptModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 72.1 | |
| SkillOptModel=GPT–5.5, Harness=Claude Code harness2026.05 | 71.5 | |
| EvoSkillModel=GPT–5.5, Harness=Claude Code harness2026.05 | 70.3 | |
| Human skillModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 66.9 | |
| Human skillModel=GPT–5.5, Harness=Claude Code harness2026.05 | 66.3 | |
| Trace2SkillModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 65.7 | |
| GEPAModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 63.9 | |
| SkillOptModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 62.8 | |
| GEPAModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 60.4 | |
| Human skillModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 59.3 | |
| No skillModel=GPT–5.5, Harness=Claude Code harness2026.05 | 57.6 | |
| SkillOptModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 56.4 | |
| LLM skillModel=GPT–5.5, Harness=Claude Code harness2026.05 | 56.4 | |
| Human skillModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 55.2 | |
| TextGradModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 54.7 | |
| GEPAModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 53.4 | |
| LLM skillModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 51.7 | |
| Trace2SkillModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 51.2 | |
| TextGradModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 51.1 | |
| SkillOptModel=GPT–5.5, Harness=Codex harness2026.05 | 51.1 | |
| No skillModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 50 | |
| SkillOptModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 50 | |
| SkillOptModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 48.8 | |
| SkillOptModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 47.1 | |
| Human skillModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 46.5 | |
| LLM skillModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 46.5 | |
| Human skillModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 45.9 | |
| No skillModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 45.9 | |
| GEPAModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 45.3 | |
| GEPAModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 45 | |
| GEPAModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 43.6 | |
| Trace2SkillModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 43 | |
| EvoSkillModel=GPT–5.5, Harness=Codex harness2026.05 | 42.4 | |
| TextGradModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 42 | |
| Human skillModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 41.9 | |
| Human skillModel=GPT–5.5, Harness=Codex harness2026.05 | 40 | |
| TextGradModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 38.7 | |
| No skillModel=GPT–5.5, Harness=Codex harness2026.05 | 38.3 | |
| LLM skillModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 36.6 | |
| No skillModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 34.9 | |
| LLM skillModel=GPT–5.5, Harness=Codex harness2026.05 | 34.3 | |
| TextGradModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 33.7 | |
| No skillModel=GPT–5.5, Harness=No harness / direct chat2026.05 | 33.1 | |
| Trace2SkillModel=Qwen3.6–35B-A3B, Harness=No harness / direct chat2026.05 | 32 | |
| TextGradModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 30 | |
| SkillOptModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 29.7 | |
| Human skillModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 22.7 | |
| GEPAModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 22.7 | |
| No skillModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 22.1 | |
| Trace2SkillModel=GPT–5.4-mini, Harness=No harness / direct chat2026.05 | 20.9 | |
| LLM skillModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 20.9 | |
| TextGradModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 20.9 | |
| LLM skillModel=GPT–5.4, Harness=No harness / direct chat2026.05 | 20.4 | |
| No skillModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 16.3 | |
| Trace2SkillModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 16.3 | |
| Trace2SkillModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 16.3 | |
| No skillModel=Qwen3.5–4B, Harness=No harness / direct chat2026.05 | 14.5 | |
| LLM skillModel=GPT–5.2, Harness=No harness / direct chat2026.05 | 14 | |
| LLM skillModel=GPT–5.4-nano, Harness=No harness / direct chat2026.05 | 12.8 |