SPADE Zero Shot - EvoBench Writing AUROC: leaderboard

The SPADE paper compares six zero-shot AI-text detectors using three released language-model backbones. AUROC and true-positive rate at 5% false positives remain separate. Paired base/instruct checkpoints are retained for Binoculars and Fast-DetectGPT. Learned soft prompts and supervised detector weights are distinct artifacts. This board reports aUROC (%). Higher is better.

Metric: AUROC (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 18 models tracked. Updated 14 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1SPADE Binoculars: Llama 3.1 8B + Llama 3.1 8B InstructEvaluation run: frozen two-checkpoint detector98.8
2SPADE Binoculars: Falcon-7B + Falcon-7B-InstructEvaluation run: frozen two-checkpoint detector98.7
3SPADE Fast-DetectGPT: Llama 3.1 8B + Llama 3.1 8B InstructEvaluation run: frozen two-checkpoint detector98.5
4SPADE Fast-DetectGPT: Falcon-7B + Falcon-7B-InstructEvaluation run: frozen two-checkpoint detector98.5
5Qwen3-8B-Base (Harness: SPADE zero-shot Likelihood)97.4
6Qwen3-8B-Base (Harness: SPADE zero-shot Entropy)96.7
7falcon-7B (Harness: SPADE zero-shot NTS)96.7
8Llama 3.1 8B (Harness: SPADE Zero Shot Likelihood)94.8
9Qwen3-8B-Base (Harness: SPADE zero-shot DNA-GPT)90.4
10falcon-7B (Harness: SPADE zero-shot DNA-GPT)90.1
11Llama 3.1 8B (Harness: SPADE Zero Shot Entropy)89.4
12falcon-7B (Harness: SPADE zero-shot Likelihood)89
13Llama 3.1 8B (Harness: SPADE Zero Shot DNA GPT)87.5
14SPADE Binoculars: Qwen3-8B-Base + Qwen3-8BEvaluation run: frozen two-checkpoint detector82.8
15Llama 3.1 8B (Harness: SPADE Zero Shot NTS)81.5

Interactive version: theaggregate.ai/benchmark?slug=spade-zero-shot-evobench-writing-auroc · How It Works · Data refreshed daily, snapshot 2026-11-16.