Frontier3DBench
Agents create editable Blender objects and scenes or executable Unreal Engine worlds across 145 tasks and nine task families. Model judgments and native project checks assess appearance, geometry, requirements, and interactions. The overall score weights object, scene, and world scores by 30%, 30%, and 40%.
Agents run through the harness named for each model at the published reasoning effort. GPT-6 Astra judges visual evidence with model identity hidden. Task assets retain individual licenses. The site publishes cost comparisons only as a raster chart and in its paper.
- Made by
- Einsia AI (Navers lab)
- Released
- November 2026
- Kind
- Spatial reasoning and counting
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Agent task
- Tasks
- All public
- Strongest model tested
- GPT-6 Astra #3 on the capability index
- Last model added
- November 2026
- Scores refreshed
- November 12th, 2026
GPT-6.1 Sol leads with 66.0
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.
- 1 GPT-6.1 Sol max 66.0
- 2 GPT-6 Astra max 63.9
- 3 Claude Opus 5.5 max 62.3
- 4
DeepSeek V4.1 Flash max 57.5 - 5 Kimi K3.1 Pro max 56.0
- 6
MiniMax-M3.1-Flash max 52.5 - 7 Kimi K3 max 49.2
- 8 GPT-6 Luna max 48.6
- 9 GLM-5.3-Flash max 47.6
- 10
Qwen 3.8 Max xhigh 46.0 - 11
Doubao Seed Evolving Nov 2026, max 42.9
Source: Official Frontier3DBench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-6.1 Sol | 66.0 | max, Codex | Sep 2026 | Official leaderboard |
| 2 | GPT-6 Astra | 63.9 | max, Codex | Sep 2026 | Official leaderboard |
| 3 | Claude Opus 5.5 | 62.3 | max, Claude Code | Sep 2026 | Official leaderboard |
| 4 | DeepSeek V4.1 Flash | 57.5 | max, DSH | Sep 2026 | Official leaderboard |
| 5 | Kimi K3.1 Pro | 56.0 | max, Kimi CLI | Oct 2026 | Official leaderboard |
| 6 | MiniMax-M3.1-Flash | 52.5 | max, Claude Code | Oct 2026 | Official leaderboard |
| 7 | Kimi K3 | 49.2 | max, Kimi CLI | Jul 2026 | Official leaderboard |
| 8 | GPT-6 Luna | 48.6 | max, Codex | Sep 2026 | Official leaderboard |
| 9 | GLM-5.3-Flash | 47.6 | max, ZCode | Aug 2026 | Official leaderboard |
| 10 | Qwen 3.8 Max | 46.0 | xhigh, Claude Code | Aug 2026 | Official leaderboard |
| 11 | Doubao Seed Evolving Nov 2026 | 42.9 | max, Claude Code | Nov 2026 | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Frontier3DBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official Frontier3DBench 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 Vision
- MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. GPT-5.1 85.4%
- 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%
- 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%
- ChartQA In use Models answer human-written and generated questions about chart images that need visual and arithmetic reasoning. Scored by relaxed accuracy with a 5 percent numeric tolerance.
Command A Vision 90.9%