Benchmark Almanac

Joke Arena

Kimi K3 1,573 leads since July 2026
Not in model cards

Models write jokes. Pairwise votes determine the published Arena rating.

The official API publishes an Arena rating, confidence interval and vote count. Voter identity requires confirmation from the board description.

Made by
Kaggle
Kind
Creative writing
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Strongest model tested
Gemini 4 Carbon #2 on the capability index
Last model added
October 2026
Scores refreshed
November 12th, 2026
25
Models scored on its own leaderboard

Kimi K3 leads with 1,573

Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.

25 models. Scroll sideways for the rest. Bars start at 1,350.

  1. 1,573 Kimi K3
  2. 1,567 Gemini 3.1 Pro
  3. 1,564 GPT-6 Astra
  4. 1,550 Claude Sonnet 5.5
  5. 1,546 Gemini 3.8 Flash
  6. 1,545 Claude Fable 5.1
  7. 1,540 GLM-5.3
  8. 1,519 GPT-5.6 Sol
  9. 1,519 GPT-6.1 Sol
  10. 1,516 GPT-5.6 Terra
  11. 1,515 Claude Opus 5
  12. 1,508 Claude Opus 5.5
  13. 1,506 GPT-5
  14. 1,500 Gemini 4 Argon
  15. 1,499 GPT-5.5
  16. 1,489 Claude Opus 4.7
  17. 1,488 GPT-5 Pro
  18. 1,486 GPT-6 Luna
  19. 1,478 Grok 4.7
  20. 1,477 GPT-5.4
  21. 1,469 Grok 4.6
  22. 1,456 Qwen 3.8 Max
  23. 1,454 GPT-5.2
  24. 1,417 Claude Sonnet 5
  25. 1,395 Gemini 4 Carbon

Bars start at 1,350.

  1. 1 Kimi K3 Kimi K3 Max Reasoning 1,573
  2. 2 Gemini 3.1 Pro Gemini 3.1 Pro Preview High Reasoning 1,567
  3. 3 GPT-6 Astra GPT-6 Astra Max Reasoning 1,564
  4. 4 Claude Sonnet 5.5 Claude Sonnet 5.5 Max Reasoning 1,550
  5. 5 Gemini 3.8 Flash Gemini 3.8 Flash High Reasoning 1,546
  6. 6 Claude Fable 5.1 Claude Fable 5.1 Max Reasoning 1,545
  7. 7 GLM-5.3 GLM-5.3 Max Reasoning 1,540
  8. 8 GPT-5.6 Sol GPT-5.6 Sol Max Reasoning 1,519
  9. 9 GPT-6.1 Sol GPT-6.1 Sol Max Reasoning 1,519
  10. 10 GPT-5.6 Terra GPT-5.6 Terra High Reasoning 1,516
  11. 11 Claude Opus 5 Claude Opus 5 Max Reasoning 1,515
  12. 12 Claude Opus 5.5 Claude Opus 5.5 Max Reasoning 1,508
  13. 13 GPT-5 GPT-5 High Reasoning 1,506
  14. 14 Gemini 4 Argon Gemini 4 Argon High Reasoning 1,500
  15. 15 GPT-5.5 GPT-5.5 High Reasoning 1,499
  16. 16 Claude Opus 4.7 Claude Opus 4.7 High Reasoning 1,489
  17. 17 GPT-5 Pro GPT-5 Pro High Reasoning 1,488
  18. 18 GPT-6 Luna GPT-6 Luna Max Reasoning 1,486
  19. 19 Grok 4.7 Grok 4.7 Xhigh Reasoning 1,478
  20. 20 GPT-5.4 GPT-5.4 Xhigh Reasoning 1,477
  21. 21 Grok 4.6 Grok 4.6 High Reasoning 1,469
  22. 22 Qwen 3.8 Max Qwen 3.8 Max Xhigh Reasoning 1,456
  23. 23 GPT-5.2 GPT-5.2 Xhigh Reasoning 1,454
  24. 24 Claude Sonnet 5 Claude Sonnet 5 High Reasoning 1,417
  25. 25 Gemini 4 Carbon Gemini 4 Carbon High Reasoning 1,395

Source: Official Joke Arena leaderboard

#ModelScoreSettingDateSource
1 Kimi K3 1,573 Kimi K3 Max Reasoning Jul 2026 Official leaderboard
2 Gemini 3.1 Pro 1,567 Gemini 3.1 Pro Preview High Reasoning Feb 2026 Official leaderboard
3 GPT-6 Astra 1,564 GPT-6 Astra Max Reasoning Sep 2026 Official leaderboard
4 Claude Sonnet 5.5 1,550 Claude Sonnet 5.5 Max Reasoning Sep 2026 Official leaderboard
5 Gemini 3.8 Flash 1,546 Gemini 3.8 Flash High Reasoning Sep 2026 Official leaderboard
6 Claude Fable 5.1 1,545 Claude Fable 5.1 Max Reasoning Sep 2026 Official leaderboard
7 GLM-5.3 1,540 GLM-5.3 Max Reasoning Aug 2026 Official leaderboard
8 GPT-5.6 Sol 1,519 GPT-5.6 Sol Max Reasoning Jul 2026 Official leaderboard
9 GPT-6.1 Sol 1,519 GPT-6.1 Sol Max Reasoning Sep 2026 Official leaderboard
10 GPT-5.6 Terra 1,516 GPT-5.6 Terra High Reasoning Jul 2026 Official leaderboard
11 Claude Opus 5 1,515 Claude Opus 5 Max Reasoning Jul 2026 Official leaderboard
12 Claude Opus 5.5 1,508 Claude Opus 5.5 Max Reasoning Sep 2026 Official leaderboard
13 GPT-5 1,506 GPT-5 High Reasoning Aug 2025 Official leaderboard
14 Gemini 4 Argon 1,500 Gemini 4 Argon High Reasoning Sep 2026 Official leaderboard
15 GPT-5.5 1,499 GPT-5.5 High Reasoning Apr 2026 Official leaderboard
16 Claude Opus 4.7 1,489 Claude Opus 4.7 High Reasoning Apr 2026 Official leaderboard
17 GPT-5 Pro 1,488 GPT-5 Pro High Reasoning Oct 2025 Official leaderboard
18 GPT-6 Luna 1,486 GPT-6 Luna Max Reasoning Sep 2026 Official leaderboard
19 Grok 4.7 1,478 Grok 4.7 Xhigh Reasoning Sep 2026 Official leaderboard
20 GPT-5.4 1,477 GPT-5.4 Xhigh Reasoning Mar 2026 Official leaderboard
21 Grok 4.6 1,469 Grok 4.6 High Reasoning Aug 2026 Official leaderboard
22 Qwen 3.8 Max 1,456 Qwen 3.8 Max Xhigh Reasoning Aug 2026 Official leaderboard
23 GPT-5.2 1,454 GPT-5.2 Xhigh Reasoning Dec 2025 Official leaderboard
24 Claude Sonnet 5 1,417 Claude Sonnet 5 High Reasoning Jun 2026 Official leaderboard
25 Gemini 4 Carbon 1,395 Gemini 4 Carbon High Reasoning Oct 2026 Official leaderboard

3 models have held number one on the official leaderboard since 2023

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

OpenAI Google DeepMind Moonshot AI
Own leaderboard 3 leaders
Kimi K3 1,573

Source: Official Joke Arena leaderboard

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
1,350 1,400 1,450 1,500 1,550 1,600 2026 2027 1,350 1,400 1,450 1,500 1,550 1,600 2026 2027

Source: Official Joke Arena leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Writing and taste

All 30 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