Benchmark Almanac

WorldCoder-Bench

GPT-5.4 27.8% leads since June 2026
Not in model cards

Agents write code that builds interactive 3D worlds. The evaluator checks whether the resulting world behaves as specified.

Made by
Shuo Lu and colleagues
Released
June 2026
Kind
Image generation and editing
Level
University
Scored
Checked automatically
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Text, 3D → Code
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-5.4 #28 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
9
Models scored in its paper

GPT-5.4 leads with 27.8% among models released by April 2026

Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 GPT-5.4 27.8%
  2. 2 Gemini 3.1 Pro 26.5%
  3. 3 Qwen 3.6 Plus 25.3%
  4. 4 MiniMax-M2.7 23%
  5. 5 DeepSeek-V3.2 21.9%
  6. 6 Claude Sonnet 4.6 18.7%
  7. 7 Kimi K2.5 18.2%
  8. 8 Claude Opus 4.6 17.5%
  9. 9 DeepSeek-V4-Flash 16.5%

Source: The WorldCoder-Bench paper

#ModelScoreSettingDateSource
1 GPT-5.4 27.8% Mar 2026 Official leaderboard
2 Gemini 3.1 Pro 26.5% Feb 2026 Official leaderboard
3 Qwen 3.6 Plus 25.3% Mar 2026 Official leaderboard
4 MiniMax-M2.7 23% Mar 2026 Official leaderboard
5 DeepSeek-V3.2 21.9% Dec 2025 Official leaderboard
6 Claude Sonnet 4.6 18.7% Feb 2026 Official leaderboard
7 Kimi K2.5 18.2% Jan 2026 Official leaderboard
8 Claude Opus 4.6 17.5% Feb 2026 Official leaderboard
9 DeepSeek-V4-Flash 16.5% Apr 2026 Official leaderboard

Number one over time

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

OpenAI
Benchmark paper Not tested since Jun 2026
GPT-5.4 27.8%

Source: The WorldCoder-Bench paper

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
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.4 DeepSeek-V3.2 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.4

Source: Official WorldCoder-Bench 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 Vision

All 636 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. Qwen 3.6 Plus 86%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy. Seed2.1 Pro 90.7%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 7 months ago, in April 2026, for DeepSeek-V4-Flash. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (9 entries).
  • Pass Models scored 9 models we track.
  • Warning Newest score The newest came 7 months ago, in April 2026, for DeepSeek-V4-Flash.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.4 (#28).
  • Pass Headroom The best is 27.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 9 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