AGORA - Finance: leaderboard
The finance and economics domain of Agora (2026), a benchmark of archive-grounded document reasoning: agents explore a large, messy collection of real workplace files with bash, locate sparse evidence, reconcile units and terminology, and compute a numeric answer. Accuracy in percent; higher is better.
Metric: Accuracy (%) on the 39 finance and economics questions (5,966 documents), archive-grounded agentic document reasoning with verifiable numeric answers over the domain collection of Markdown workplace documents, agent in the mini-swe-agent harness (bash tool only), temperature 1.0 with reasoning effort at the maximum each model supports; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.1 | 56.41 | #179 |
| 2 | GPT-5.5 (xHigh) | 53.85 | #49 (GPT-5.5) |
| 3 | DeepSeek V4 Pro (Max) | 46.15 | #188 (DeepSeek V4 Pro) |
| 4 | Gemini 3.1 Pro (Preview) (High) | 41.03 | #94 (Gemini 3.1 Pro (Preview)) |
| 5 | DeepSeek V4 Flash (Max) | 38.46 | #258 (DeepSeek V4 Flash) |
| 6 | Gemini 3.1 Flash Lite (High) | 5.13 | #364 (Gemini 3.1 Flash Lite) |
| 7 | Qwen 3.5 35B A3B | 2.56 | #482 |
| 8 | Qwen 3.5 9B | 0 | #803 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=agora-finance · How It Works · Data refreshed daily, snapshot 2026-11-13.