ESF-Bench - Tagged Slot Accuracy: leaderboard

Metric: Tagged slot accuracy (%; accuracy over the 66.8% of slots tagged with at least one taxonomy scenario; 810 synthetic multi-turn enterprise slot-filling samples (6,530 slots, 8 domains) whose prompts combine a schema, a user-assistant conversation and auxiliary sources; varied output formats mapped to a standard form before exact comparison with the ground truth; 32k maximum tokens, reasoning setting as printed per model). Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.

Top models

#ModelScore
1Gemini 2.5 Flash80.9
2GPT-5.1 (High)78.2
3GPT-OSS-120B (High)76.5
4Qwen 3 32B (Thinking)75.3
5Ministral-3-14B-Reasoning-251274.1
6GPT-OSS-120B (Low)73
7GPT-5.1 (Non-reasoning)67.3
8Gemini 2.5 Flash (Non-reasoning)64.5
9Ministral-3-14B-Instruct-251256.7
10Qwen 3 32B (Non-reasoning)56.3

Interactive version: theaggregate.ai/benchmark?slug=esf-bench-tagged-slot-accuracy · How It Works · Data refreshed daily, snapshot 2026-09-29.