SABRE-Prior - Context: leaderboard

Metric: Strict All4 accuracy (%; a case counts only when all four yes/no probes on the base and edited image are right: an expected entity replaced by a contextually unexpected one; zero-shot, greedy decoding where supported, 8-16 output tokens for API models; 100 generated or edited cases per subset that survived filtering against Gemini 3.5 Flash and human review). Source: arxiv.org. Saturation forecast: Around 2029. 6 models tracked.

Top models

#ModelScore
1Claude Sonnet 4.610
2Kimi K2.6 (Non-reasoning)7
3Grok 4.34
4Qwen 3.5 27B (Non-reasoning)3
5GPT-5.4 (Non-reasoning)1

Interactive version: theaggregate.ai/benchmark?slug=sabre-prior-context · How It Works · Data refreshed daily, snapshot 2026-09-29.