GTA-Workflow - Operation: leaderboard

Root success on the GTA-Workflow tasks that require operation tools, part of GTA-Workflow, the open-ended tier of GTA-2 (2026): 132 long-horizon productivity workflows from real user queries with deployed tools and multimodal inputs, run end to end in a ReAct agent framework and judged by GPT-5.2 against recursive checkpoint rubrics. Percent; higher is better.

Metric: Root success rate (%) on the tasks that require operation tools on GTA-Workflow, the open-ended long-horizon workflow tier of GTA-2 (132 real-user tasks with deployed tools and multimodal inputs), run end to end in the default Lagent ReAct framework; deliverables judged by GPT-5.2 against recursive checkpoint rubrics; higher is better. Source: arxiv.org. Saturation forecast: Around January 2028. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro13.1#276
2GPT-510.71#178
3Grok 410.71#319
4Qwen 3 235B A22B9.52#628
5Llama 4 Scout9.52#1818
6Claude Sonnet 4.58.33#261
7DeepSeek V3.27.14#419
8Kimi K25.95#443
9Llama 3.1 70B Instruct1.19#1571
10Llama 3.2 3B Instruct1.19#7241
11Qwen 3 30B A3B1.19#1156
12Llama 3.1 8B Instruct0#4600
13Qwen 3 8B0#1873

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=gta-workflow-operation · How It Works · Data refreshed daily, snapshot 2026-11-21.