MARCA (Portuguese, Orchestrator): leaderboard
Metric: Checklist accuracy (0-1, times 100): share of a question's checklist items (expected entities and attributes) that a GPT-4.1 judge marks satisfied, averaged over MARCA's 52 Portuguese questions and three runs, Orchestrator framework (the model delegates sub-questions to subagents that call web_search and web_scrape); higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro | 88.9 |
| 2 | GPT-5.2 | 87.2 |
| 3 | GPT-4.1 | 82.5 |
| 4 | Gemini 3 Flash | 82 |
| 5 | GPT-5 Mini | 81.9 |
| 6 | Kimi K2 | 77.3 |
| 7 | GPT-4.1 Mini | 73.5 |
| 8 | Qwen 3 235B A22B | 67.8 |
| 9 | Gemini 2.5 Pro | 59.8 |
| 10 | Gemini 2.5 Flash | 51.7 |
| 11 | Qwen 3 30B A3B | 49.3 |
| 12 | Gemini 2.5 Flash Lite | 35.4 |
Interactive version: theaggregate.ai/benchmark?slug=marca-portuguese-orchestrator · How It Works · Data refreshed daily, snapshot 2026-10-07.