TFRBench - Solar Daily - Domain Relevance: leaderboard

Forecasting reasoning on the daily solar power series of TFRBench (2026): the model reasons step by step about trends and external events before forecasting, and Gemini 3 Pro compares that reasoning with verified reference reasoning. Mean judge score from 1 to 5 for domain relevance, the use of domain-specific terminology and logic; higher is better.

Metric: Mean LLM-judge score (1-5) for domain relevance of the model's chain-of-thought forecasting reasoning on the TFRBench daily solar power (GIFT-Eval, energy) windows, judged by Gemini 3 Pro against the benchmark's verified reference reasoning (w/ Reasoning setting: the model reasons step by step from the history window, then forecasts; no tools; mean of three runs); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 5 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Sonnet 4.54.92#261
2Gemini 3 Pro4.7#145
3Claude Sonnet 44.5#410
4Gemini 2.5 Pro3.03#276
5Gemini 2.5 Flash2.47#463

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

Interactive version: theaggregate.ai/benchmark?slug=tfrbench-solar-daily-domain-relevance · How It Works · Data refreshed daily, snapshot 2026-11-21.