PinpointQA - Target Presence Verification: leaderboard

Target presence verification on PinpointQA (2026): from 64 frames of an indoor ScanNet++ or ScanNet200 video, the model answers yes or no to whether a named small object appears. Exact-match accuracy scaled to 0-100; higher is better.

Metric: Exact-match accuracy (times 100, micro over QA pairs) of yes/no target presence verification for small objects of the PinpointQA evaluation split (2,019 QA pairs from held-out ScanNet++ and ScanNet200 scenes; 64 uniformly sampled RGB frames per indoor video, no depth or 3D input, greedy decoding); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Kimi K2.580#266
2Qwen 3 VL 8B Instruct78#1029
3LLaVA-OneVision-1.5-8B76#1864
4Cambrian-S-7B73#2198
5GPT-5.465#135
6InternVL3.5-8B-Instruct65#1891
7SenseNova-SI-1.3-InternVL3-8B64#1610
8Spatial-MLLM-820K-4B52

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

Interactive version: theaggregate.ai/benchmark?slug=pinpointqa-target-presence-verification · How It Works · Data refreshed daily, snapshot 2026-11-21.