InternVL3.5-14B-Instruct: benchmark results

Provider: Shanghai AI Lab. Access: Open.

Unified ELO 1530 ± 25, rank #651 of 1605 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SiT-Bench - Global Perception & Mapping23.64Accuracy (%; item-weighted over the category's subtasks)76.7
SiT-Bench - Logic & Anomaly Detection37.56Accuracy (%; item-weighted over the category's subtasks)76.7
SiT-Bench - Navigation & Planning49.33Accuracy (%; item-weighted over the category's subtasks)76.7
MCSBench - L326.72Accuracy (%)76.3
SiT-Bench41.47Accuracy (%; item-weighted over the 17 subtasks)69
SiT-Bench - Embodied & Fine-grained Perception42.81Accuracy (%; item-weighted over the category's subtasks)63.3
SiT-Bench - Multi-View & Geometric Reasoning46.17Accuracy (%; item-weighted over the category's subtasks)61.7
MCSBench - L253.27Accuracy (%)60.5
MCSBench - Overall48.41Accuracy (%)59.2
MCSBench - L156.32Accuracy (%)48
K-MetBench47.9Accuracy (self-reported)32.8
Video-IFBench - Selection12.5Task-gated instruction satisfaction rate (%; TISR on selecti29.7

Interactive version: theaggregate.ai/model?slug=internvl3-5-14b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-26.