Benchmark Almanac

Slop Score

human-baseline 10.4 tops its leaderboard Kimi K2 Sep 2025 18.3 leads the models we track Not in model cards

Model writing is scored by how often it uses words and patterns that are far more common in AI text than in human text. Lower means less slop.

The maker says it is not an AI detector. Release month is the repo creation month.

Made by EQ-Bench (Sam Paech) Released Oct 2025 Taste and style All tasks public Last model added Sep 2025 Scores refreshed 2 Nov 2026
Level
Everyday Most adults, without training
Scored
Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers
Open-ended Free-form text, documents, designs or media judged as a whole.
sam-paech/slop-score 42 stars Leaderboard eqbench.com
22
Entries on its own leaderboard, 17 models we track

human-baseline tops the leaderboard with 10.4

Best score per entry from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Kimi K2, at 18.3.

The newest model with a score here came out in Sep 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 human-baseline 10.4
  2. 2 Kimi K2 Sep 2025 18.3
  3. 3 Claude Sonnet 4.5 19.5
  4. 4 sam-paech/gemma-3-27b-it-antislop 20.8
  5. 5 GPT-5 mini 26.3
  6. 6 Claude Sonnet 4 27.3
  7. 7 openrouter/polaris-alpha 30.8
  8. 8 o3 31.8
  9. 9 GPT-5 nano 34.1
  10. 10 openai/chatgpt-4o-latest 47.7
  11. 11 Llama 4 Maverick 49.2
  12. 12 Mistral Small 3.2 52.7
  13. 13 DeepSeek-V3.2-Exp 54.2
  14. 14 Mistral NeMo 54.4
  15. 15 inclusionai/ling-1t 56.2
  16. 16 GLM-4.5 59.9
  17. 17 Qwen3-4B 62.2
  18. 18 DeepSeek-R1 May 2025 67.9
  19. 19 Gemma 3 27B 69.5
  20. 20 Gemma 3 12B 70.6
  21. 21 Gemini 2.5 Flash Sep 2025 77.7
  22. 22 Gemma 3 4B 85.2
#EntryScoreSettingDateSource
1 human-baseline 10.4 Official leaderboard
2 Kimi K2 Sep 2025 18.3 Sep 2025 Official leaderboard
3 Claude Sonnet 4.5 19.5 Sep 2025 Official leaderboard
4 sam-paech/gemma-3-27b-it-antislop 20.8 Official leaderboard
5 GPT-5 mini 26.3 Aug 2025 Official leaderboard
6 Claude Sonnet 4 27.3 May 2025 Official leaderboard
7 openrouter/polaris-alpha 30.8 Official leaderboard
8 o3 31.8 Apr 2025 Official leaderboard
9 GPT-5 nano 34.1 Aug 2025 Official leaderboard
10 openai/chatgpt-4o-latest 47.7 Official leaderboard
11 Llama 4 Maverick 49.2 Apr 2025 Official leaderboard
12 Mistral Small 3.2 52.7 Jun 2025 Official leaderboard
13 DeepSeek-V3.2-Exp 54.2 Sep 2025 Official leaderboard
14 Mistral NeMo 54.4 Jul 2024 Official leaderboard
15 inclusionai/ling-1t 56.2 Official leaderboard
16 GLM-4.5 59.9 Aug 2025 Official leaderboard
17 Qwen3-4B 62.2 Apr 2025 Official leaderboard
18 DeepSeek-R1 May 2025 67.9 May 2025 Official leaderboard
19 Gemma 3 27B 69.5 Mar 2025 Official leaderboard
20 Gemma 3 12B 70.6 Mar 2025 Official leaderboard
21 Gemini 2.5 Flash Sep 2025 77.7 Sep 2025 Official leaderboard
22 Gemma 3 4B 85.2 Mar 2025 Official leaderboard

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Moonshot AI
Own leaderboard Not tested since Oct 2025
Kimi K2 18.3

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
0.00 20.0 40.0 60.0 80.0 100.0 2025 0.00 20.0 40.0 60.0 80.0 100.0 2025

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Writing and taste

All 24 in Writing and taste
  • Design Bench (Moonshot) In use Models make graphic designs from direct prompts or few-shot reference images. An automated grader rates the results as Elo scores. Kimi K3.1 Pro 957
  • WritingBench Fading Models complete 1,000 real-world writing requests across 6 domains and 100 subdomains. A judge model scores each reply on five query-specific criteria. GPT-5 83.9
  • Creative Writing (MiniMax in-house) Only in system cards Models write creative pieces for real Hailuo AI user requests. Automatic assessments score the writing. MiniMax-Text-01 81.3%
  • DeepSeek Chinese functional writing evaluation Only in system cards Models answer 3,170 everyday Chinese writing prompts such as reports, emails and speeches. Pairwise judging against a baseline model gives a win rate. DeepSeek-V4-Pro 62.6%
  • Slide generation human evaluation (Z.ai) Only in system cards People compare HTML slide decks from two models on content, layout and visual quality. Metric is overall win rate. GLM-5 67.5%
  • Task-dependent output homogenization evaluation Only in system cards Models answer prompts several times. A judge scores functional diversity where it is wanted and consistency where one answer is right. Muse Spark 0.21