AA-3D
Interactive 3D experiences built from a prompt, graded head to head, so the score is an Elo rather than a percentage.
As of 18 Oct 2026, Claude Sonnet 5.5 leads AA-3D on BenchLeader with 1197, ahead of Claude Opus 5.5 at 1191, across 21 model configurations with a published result.
- Published by
- Artificial Analysis
- Category
- Multimodal
- Index weight
- Reference only
- Models
- 21
- Data as of
- 18 Oct 2026
Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.
- 1Claude Sonnet 5.5 (max)1197
- 2Claude Opus 5.5 (max)1191
- 3GPT-6 Astra (max)1166
- 4Gemini 4 Argon (high)1154
- 5Claude Fable 5.1 (max)1133
- 6GPT-6 Sol (max)1123
- 7GPT-6.1 Sol (max)1118
- 8Muse Spark 1.3 (max)1037
- 9GPT-6 Luna (max)1000
- 10Claude Haiku 5.5 (max)997
- 11Grok 4.7 (xhigh)995
- 12Kimi K3 (max)981
- 13MiMo-V2.6-Pro980
- 14Gemini 3.8 Flash (high)969
- 15Qwen3.8 Max (max)948
21 of 21
| # | ||||
|---|---|---|---|---|
| 1 |  Claude Sonnet 5.5max | 1197 | 67.5 | 2026-09-28 |
| 2 | Claude Opus 5.5max | 1191 | 71.9 | 2026-09-22 |
| 3 |  GPT-6 Astramax | 1166 | 69.9 | 2026-09-03 |
| 4 |  Gemini 4 Argonhigh | 1154 | 71.4 | 2026-09-30 |
| 5 | Claude Fable 5.1max | 1133 | 70.1 | 2026-09-01 |
| 6 | GPT-6 Solmax | 1123 | 65.8 | 2026-09-22 |
| 7 | GPT-6.1 Solmax | 1118 | 69.3 | 2026-09-29 |
| 8 |  Muse Spark 1.3max | 1037 | 67.4 | 2026-09-02 |
| 9 | GPT-6 Lunamax | 1000 | 58.7 | 2026-09-22 |
| 10 | Claude Haiku 5.5max | 997 | 57.9 | 2026-10-07 |
| 11 |  Grok 4.7xhigh | 995 | 60.7 | 2026-09-21 |
| 12 |  Kimi K3maxopen ↗ | 981 | 65.1 | 2026-07-16 |
| 13 |  MiMo-V2.6-Proopen ↗ | 980 | 65.6 | 2026-09-21 |
| 14 | Gemini 3.8 Flashhigh | 969 | 63.5 | 2026-09-02 |
| 15 |  Qwen3.8 Maxmax | 948 | 63.9 | 2026-09-02 |
| 16 |  GLM 5.3 Flashopen ↗ | 935 | 62.4 | 2026-08-26 |
| 17 |  Mistral Large 4open | 870 | 59.0 | 2026-10-06 |
| 18 |  Step 5 Previewopen | 852 | 62.4 | 2026-09-18 |
| 19 | GLM 5.3maxopen ↗ | 787 | 63.8 | 2026-08-18 |
| 20 |  MiniMax M3open ↗ | 740 | 56.4 | 2026-06-01 |
| 21 |  Nemotron 3 Ultra 550B A55Bthinkingopen ↗ | 569 | 55.9 | 2026-06-04 |
More multimodal benchmarks · All LLM benchmarks
Cite as: BenchLeader, “AA-3D leaderboard”, https://www.benchleader.com/benchmarks/aa_3d, data as of 18 Oct 2026.
AA-3D: questions
- What does AA-3D measure?
- Interactive 3D experiences built from a prompt, graded head to head, so the score is an Elo rather than a percentage. Scores are reported in Elo-style ratings from pairwise comparisons; higher is better.
- Which AI model leads AA-3D?
- Claude Sonnet 5.5 leads AA-3D with 1197 as of 18 Oct 2026, ahead of Claude Opus 5.5 at 1191.
- How many models have AA-3D results?
- 21 model configurations have a AA-3D result on BenchLeader, all taken from Artificial Analysis.
- Who runs AA-3D and how often is it updated?
- AA-3D is published by Artificial Analysis. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does AA-3D count toward the BenchLeader Index?
- No. AA-3D is shown for reference but left out of the composite index.