EgoPoint-Bench - Simulation Basic Perception: leaderboard

EgoPoint-Bench (2026) asks multimodal models about the object a first-person hand points at, on held-out simulation scenes and real smart-glasses recordings, zero-shot; true/false and single-choice answers by exact match and open answers scored by a GPT-4o judge. This board covers the Basic Perception questions (category, colour and texture) among the 1,929 simulation test questions. Accuracy in percent; higher is better.

Metric: Accuracy (%) on the Basic Perception questions among the 1,929 held-out simulation test questions (HM3D, HSSD and AI2-THOR scenes); zero-shot (direct inference); true/false and single-choice questions by exact match, open-ended answers scored 0 to 1 by a GPT-4o judge; greedy decoding for open models, temperature 0 for API models; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-5 Mini59.96#364
2Qwen 3 VL 8B Instruct57.55#1029
3Qwen 3 VL 32B Instruct56.52#606
4GLM-4.6V-Flash56.16#1040
5Gemini 3 Flash54.39#175
6GPT-5.2 Instant54.14#304
7InternVL3.5-8B-Instruct52.86#1891
8Gemini 3 Pro52.47#145
9InternVL3.5-2B-Instruct51.97#6517
10LLaVA-v1.5-7B50.83#7630
11LLaVA-NeXT-7B47.42#5032
12InternVL3.5-14B-Instruct46.79#1736

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=egopoint-bench-simulation-basic-perception · How It Works · Data refreshed daily, snapshot 2026-11-21.