STQA - Human-Authored Questions - Forecast MAE: leaderboard
STQA tests stock-focused tabular question answering over historical and forecast data. QR is historical query reasoning, NF numerical forecasting and FR forecast-based reasoning. The SQFRS tool workflow and vanilla prompting are separate configurations. This series reports STQA - Human-Authored Questions - Forecast MAE from 2609.06117v1, S5.T4: Mean Absolute Error. Lower is better.
Metric: Mean Absolute Error. Source: arxiv.org. Saturation forecast: Around April 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-OSS-120B | 2.76 | #731 |
| 2 | GPT-OSS-20B | 2.77 | #1207 |
| 3 | Qwen 3 30B A3B | 2.8 | #1196 |
| 4 | Gemma 4 31B | 2.83 | #392 |
No result here: #4 Claude Opus 5.5, #5 GPT-6 Astra, #9 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=stqa-human-authored-questions-forecast-mae · How It Works · Data refreshed daily, snapshot 2026-11-08.