Benchmark Almanac

ScanNet-1500

MultiLoc-518 55.6% tops its leaderboard
Not in model cards

Models match features between pairs of indoor views and estimate relative camera poses. The evaluation measures pose accuracy on 1,500 image pairs.

Meta compares areas under the pose error curve at two angular thresholds.

Released
November 2019
Kind
Spatial reasoning and counting
Level
School
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Images → 3D
Interaction
Single turn
Items from
Another benchmark
Refresh
One fixed set
Run by
Anyone
Based on
ScanNet
16
Entries on SOTA2’s leaderboard, 0 models we track

MultiLoc-518 tops the leaderboard with 55.6%

Best score per entry from SOTA2’s leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

  1. 1 MultiLoc-518 Approach Type=PR, Input Resolution=518, Zero-shot generalization=true 55.6%
  2. 2 VGGT-518 Approach Type=PR, Input Resolution=518, Protocol=Independent Benchmarking 38.9%
  3. 3 MASt3R+Ours distillation=integrated 37.9%
  4. 4 MASt3R distillation=original 37.6%
  5. 5 VGGT+Ours distillation=integrated 35.2%
  6. 6 Reloc3r 34.8%
  7. 7 ReLoc3r-512 Approach Type=PR, Input Resolution=512, Protocol=Independent Benchmarking 34.8%
  8. 8 DUSt3R+Ours distillation=integrated 33.7%
  9. 9 NoPoSplat 31.8%
  10. 10 DUSt3R distillation=original 31.6%
  11. 11 ROMA 28.9%
  12. 12 VGGT distillation=original 28.4%
  13. 13 ReLoc3r-224 Approach Type=PR, Input Resolution=224, Protocol=Independent Benchmarking 28.3%
  14. 14 Efficient LoFTR 19.2%
  15. 15 ExReNet Approach Type=PR, Supervision=SN 2.30%
  16. 16 Map-free Approach Type=PR, Supervision=Regress-SN 1.84%

Source: SOTA2’s leaderboard

#EntryScoreSettingDateLinkSource
1 MultiLoc-518 55.6% Approach Type=PR, Input Resolution=518, Zero-shot generalization=true sota2.com Official leaderboard
2 VGGT-518 38.9% Approach Type=PR, Input Resolution=518, Protocol=Independent Benchmarking sota2.com Official leaderboard
3 MASt3R+Ours 37.9% distillation=integrated sota2.com Official leaderboard
4 MASt3R 37.6% distillation=original sota2.com Official leaderboard
5 VGGT+Ours 35.2% distillation=integrated sota2.com Official leaderboard
6 Reloc3r 34.8% sota2.com Official leaderboard
7 ReLoc3r-512 34.8% Approach Type=PR, Input Resolution=512, Protocol=Independent Benchmarking sota2.com Official leaderboard
8 DUSt3R+Ours 33.7% distillation=integrated sota2.com Official leaderboard
9 NoPoSplat 31.8% sota2.com Official leaderboard
10 DUSt3R 31.6% distillation=original sota2.com Official leaderboard
11 ROMA 28.9% sota2.com Official leaderboard
12 VGGT 28.4% distillation=original sota2.com Official leaderboard
13 ReLoc3r-224 28.3% Approach Type=PR, Input Resolution=224, Protocol=Independent Benchmarking sota2.com Official leaderboard
14 Efficient LoFTR 19.2% sota2.com Official leaderboard
15 ExReNet 2.30% Approach Type=PR, Supervision=SN sota2.com Official leaderboard
16 Map-free 1.84% Approach Type=PR, Supervision=Regress-SN sota2.com Official 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

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the official leaderboard (16 entries).
  • Pass Models scored 16 entries on its leaderboard, none a model we track.
  • Note Newest score No scores yet.
  • Note Frontier models No models scored yet.
  • Note Headroom No scores yet.
Leaderboard import 3 pass
  • Pass Importer We read the board on SOTA2: 16 entries at the last read, none a model we track.
  • 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