ARFBench - Identification: leaderboard

Time series question answering on software incident data in ARFBench (2026, Datadog): multiple-choice questions over multivariate production telemetry from real incidents, with the series shown as chart images. Accuracy on the Identification questions (which of three channels are anomalous), in percent; higher is better.

Metric: Accuracy (%) on the 38 Identification questions (which of three channels are anomalous) of ARFBench, for the time series rendered as chart images (Matplotlib plots, three images for paired questions), with the series description and channel names as text; few-shot multiple-choice prompting with shuffled answer choices, temperature 0.05 (default for models without the setting), reasoning models at medium effort with 2,000 output tokens, blank or invalid answers counted wrong; higher is better. Source: arxiv.org. Saturation forecast: Around 2034. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Toto-1.0-QA-Experimental 32B (TSFM-VLM)47.4
2GPT-4o39.5#751
3Qwen 3 VL 32B (post-trained)36.8
4GPT-531.6#178
5Claude Opus 4.631.6#113
6GPT-5.4Medium31.6#135
7GPT-4.128.9#505
8Gemini 3 Pro28.9#145
9Qwen 3 VL 8B26.3#917
10Qwen 3 VL 32B23.7#622
11Claude Sonnet 4.518.4#261

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

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