Leaderboard FUNGI

One or more photos reconstructed into a 3D mesh.

Ranked by human votes · 1614 cast · see the AI-judge board →

Share Post on X
How ranking works

Rank groups models whose 95% bootstrap CIs overlap into the same rank (they are not statistically separable). BT = Bradley–Terry score (Elo-scaled); the bar shows the 95% CI with the point estimate marked. Every board ranks a SINGLE method — scores from different methods come from disconnected match pools and aren't comparable, so there is no cross-method ranking. Elo updates live per vote. Recompute in Admin.

Image→3D reconstruction

10 generators
Rank Generator BT score 95% interval BT 275.7–1535.3 Trend Votes Status
1 TRELLIS 2 papercode 1246.8
[1101.0, 1535.3]
17 13 more votes → firm
1 Meshy 6 papercode 1102.6
[993.6, 1299.5]
27 3 more votes → firm
1 Hunyuan3D v3 papercode 1068.3
[993.5, 1160.7]
36 firm
2 Hunyuan3D v2 papercode 1016.2
[922.2, 1085.8]
31 firm
2 SAM 3D papercode 984.5
[838.3, 1090.1]
27 3 more votes → firm
2 Pixal3D papercode 980.4
[860.1, 1077.4]
31 firm
2 Hunyuan3D 3.1 papercode 952.4
[790.4, 1067.8]
27 3 more votes → firm
1 Rodin/Hyper3D papercode 915.7
[682.3, 1109.2]
17 13 more votes → firm
4 TripoSR papercode 840.3
[595.1, 985.9]
20 10 more votes → firm
5 TRELLIS via Replicate papercode 755.9
[468.1, 899.5]
17 13 more votes → firm
TRELLIS via fal papercode 706.8
[275.7, 904.8]
10 20 more votes → firm

95% credible interval · point estimate · trend · firm a rank backed by 30+ votes; below that, Status counts the votes still needed · click a row for detail

Filters & bias audit

Bias audit — left(A) win rate 0.5 (≈0.50 = unbiased) · tie 0.07 · bad 0.162

VLM judge (Sonnet 4.6, multi-view) — automated LLM-judge rankings by paradigm

Loading automated rankings…