Benchmark Almanac

ASCII Bench (Jason Botterill)

Gemini 4 Argon 1,948 leads since September 2026
Not in model cards

Models draw pictures using plain text characters. Readers compare two anonymous drawings for the same prompt. The leaderboard estimates ratings from their votes.

The maker rates six models from eight votes. Seven additional models have drawings but no rating. Readers can choose a winner or give both models equal credit.

Made by
Jason Botterill
Kind
Creative writing
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Measures
Preference
Subject
Arts
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Live play or votes
Refresh
Live
Run by
A host
In views
Kept fresh
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
13
Models scored on its own leaderboard

Gemini 4 Argon leads with 1,948

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.

Bars start at 1,200.

  1. 1 Gemini 4 Argon high 1,948
  2. 2 Claude Opus 5.5 high 1,730
  3. 3 GPT-6.1 Sol high 1,696
  4. 4 Gemini 3.8 Flash high 1,546
  5. 5 Claude Sonnet 5.5 high 1,539
  6. 6 GLM-5.3 high 1,493
  7. 7 Kimi K3.1 Pro high 1,418
  8. 8 DeepSeek V4.1 Flash high 1,417
  9. 9 Grok 4.7 high 1,396
  10. 10 Claude Haiku 5.5 high 1,395
  11. 11 MiniMax-M3.1-Flash high 1,384
  12. 12 Qwen 3.8 Max high 1,368
  13. 13 GPT-6 Luna high 1,338

Source: Official ASCII Bench (Jason Botterill) leaderboard

#ModelScore ±SettingDateLinkSource
1 Gemini 4 Argon 1,948 1857 to 2070 high Sep 2026 asciibench.com Official leaderboard
2 Claude Opus 5.5 1,730 1657 to 1811 high Sep 2026 asciibench.com Official leaderboard
3 GPT-6.1 Sol 1,696 1646 to 1762 high Sep 2026 asciibench.com Official leaderboard
4 Gemini 3.8 Flash 1,546 ±55 high Sep 2026 asciibench.com Official leaderboard
5 Claude Sonnet 5.5 1,539 1476 to 1595 high Sep 2026 asciibench.com Official leaderboard
6 GLM-5.3 1,493 1438 to 1552 high Aug 2026 asciibench.com Official leaderboard
7 Kimi K3.1 Pro 1,418 1359 to 1469 high Oct 2026 asciibench.com Official leaderboard
8 DeepSeek V4.1 Flash 1,417 1362 to 1478 high Sep 2026 asciibench.com Official leaderboard
9 Grok 4.7 1,396 1345 to 1454 high Sep 2026 asciibench.com Official leaderboard
10 Claude Haiku 5.5 1,395 1350 to 1447 high Oct 2026 asciibench.com Official leaderboard
11 MiniMax-M3.1-Flash 1,384 1331 to 1430 high Oct 2026 asciibench.com Official leaderboard
12 Qwen 3.8 Max 1,368 1303 to 1420 high Aug 2026 asciibench.com Official leaderboard
13 GPT-6 Luna 1,338 1286 to 1386 high Sep 2026 asciibench.com Official leaderboard

5 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.

Qwen Z.ai Google DeepMind Anthropic
Own leaderboard 5 leaders

Source: Official ASCII Bench (Jason Botterill) leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard
1,200 1,400 1,600 1,800 2,000 2027 1,200 1,400 1,600 1,800 2,000 2027

Source: Official ASCII Bench (Jason Botterill) leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Writing and taste

All 65 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

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (13 entries).
  • Pass Models scored 13 models we track.
  • Pass Newest score The newest is from October 2026, for Kimi K3.1 Pro.
  • Pass Frontier models 4 of the top 10 have a score, led by Claude Opus 5.5 (#5).
  • Note Headroom Scored as an Elo rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on asciibench.com: 13 rows for models we track at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check