VoxelBench Image-Prompt Leaderboard (VoxelBench Image)

We show this table for reference; we do not rank on it.

Data verified September 29, 202636 confirmed releases in the last 30 daysFollow model changes

A live human-preference benchmark where multimodal models build voxel structures from image references and voters compare anonymous results produced from the same prompt.

Glicko-2 rating on VoxelBench image-prompt leaderboard — September 29, 2026

We mirror the published glicko-2 rating view for VoxelBench image-prompt leaderboard. Gemini 3.1 Pro Preview leads the public snapshot at 1642, followed by GPT-5.4 (1490) and Gemini 3 Pro Preview (1466). We do not use these results to rank models overall.

Gemini 3.1 Pro Preview

Google

2,230 votes · 81.1% win · 95% CI 1538–1746

1642

2Proprietary

GPT-5.4

OpenAI

2,938 votes · 64.9% win · 95% CI 1404–1576

1490

Overall 68.22Context 1.05M

Gemini 3 Pro Preview

Google

5,167 votes · 66.6% win · 95% CI 1388–1544

1466

12 modelsMultimodal & GroundedCurrentDisplay onlyUpdated September 29, 2026

Glicko-2 rating table (12 models)

Score

Gemini 3.1 Pro PreviewGoogle2,230 votes · 81.1% win · 95% CI 1538–1746

1642

GPT-5.4OpenAI · Closed2,938 votes · 64.9% win · 95% CI 1404–1576

1490

Gemini 3 Pro PreviewGoogle5,167 votes · 66.6% win · 95% CI 1388–1544

1466

Claude Opus 4.6Anthropic · Closed4,160 votes · 65.6% win · 95% CI 1357–1503

1430

GPT-5.2 (xHigh)OpenAI5,038 votes · 64.1% win · 95% CI 1330–1478

1404

Claude Opus 4.5Anthropic · Closed5,075 votes · 55.5% win · 95% CI 1202–1350

1276

Claude Opus 4.7Anthropic · Closed4,441 votes · 49.0% win · 95% CI 1167–1313

1240

GPT-5 (high)OpenAI · Closed4,000 votes · 46.2% win · 95% CI 1143–1303

1223

Gemini 2.5 ProGoogle · Closed3,804 votes · 43.9% win · 95% CI 1139–1299

1219

GPT-5 Codex (High)OpenAI4,010 votes · 43.4% win · 95% CI 1122–1278

1200

Claude Opus 4.1 (64K Thinking)Anthropic4,025 votes · 38.0% win · 95% CI 1069–1225

1147

GPT-5.1OpenAI · Closed5,159 votes · 39.0% win · 95% CI 1071–1219

1145

How to read this leaderboard

The board measures which rendered voxel builds voters prefer when models receive an image reference. Higher ratings indicate stronger pairwise results against the tested field, not absolute reconstruction accuracy.

Operator receipt: 12 sourced rows are currently displayable on this page; the leading published row is Gemini 3.1 Pro Preview at 1642.

Honest limit: The source does not publish a fixed, versioned task set or a controlled reconstruction metric. Ratings can move as prompts, models, and voters change, and the complete generation pipeline contributes to every result.

VoxelBench text-prompt leaderboard Multimodal and grounded rankings

Official live leaderboard Blind comparison voting

How we show VoxelBench image-prompt ratings

We mirror the image-prompt table from VoxelBench's public leaderboard API. The source page shows models after at least 50 votes and ranks the eligible rows with Glicko-2; the snapshot keeps each rating, deviation, 95% confidence interval, vote count, win rate, and win/loss/tie record.

Voters compare two anonymous voxel structures produced for the same image-reference prompt. The result mixes model behavior with VoxelBench's generation tools, instructions, settings, prompt mix, renderer, and voter pool, so we keep it display only and separate from weighted model rankings.

Snapshot

12 eligible model rows50+ votes per modelImage-prompt buildsBlind pairwise votesGlicko-2Display only

VoxelBench leaderboard Blind comparison voting VoxelBench prompts

The published VoxelBench Image snapshot places Gemini 3.1 Pro Preview first at 1642. The third row is 176 score units behind. The broader top-10 range is 442 score units, so the table still separates the published systems.

12 models have been evaluated on VoxelBench Image. The benchmark falls in the Multimodal & Grounded category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. VoxelBench Image is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About VoxelBench Image

Year

2025

Tasks

Live image-reference prompts for 3D voxel construction

Format

Glicko-2 rating from blind pairwise votes

Difficulty

Visual grounding, 3D construction, and aesthetic quality

The mirrored image-prompt view keeps only models with at least 50 votes, matching the public page. Each row preserves the Glicko-2 rating, deviation, confidence interval, vote count, win rate, and win/loss/tie record.

VoxelBench leaderboard Public benchmark source

Freshness and provenance

Version

Live VoxelBench Glicko-2

Refresh cadence

Rolling

Staleness state

Current

Question availability

Prompts browsable; full task set not versioned

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does VoxelBench Image measure?

A live human-preference benchmark where multimodal models build voxel structures from image references and voters compare anonymous results produced from the same prompt.

Which model leads the published VoxelBench Image snapshot?

Gemini 3.1 Pro Preview currently leads the published VoxelBench Image snapshot with 1642 glicko-2 rating. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on VoxelBench Image?

The September 29, 2026 snapshot contains 12 AI models.

Last updated: September 29, 2026 · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 2,000+ readers.

One email each week. Unsubscribe anytime.