PRISM-VLM - Quality: leaderboard

Metric: Per-axis mean score (%; base accuracy on the recycled question, judged against the source gold answer; over PRISM-VLM's 6,238 items recycled from 15 public VLM benchmarks (five seeds of 100 items per benchmark); GPT-5 (low effort) synthesizes the perturbations and grades the open-ended axes; each model at the lowest reasoning effort its provider exposes, temperature 0). Source: arxiv.org. Saturation forecast: Around 2029. 42 models tracked.

Top models

#ModelScore
1Gemini 3 Flash (Minimal)74.1
2Qwen 3.5 9B (Non-reasoning)69.9
3GPT-4.1 Mini69.7
4GPT-5.4 Mini69.2
5Gemini 2.5 Flash (Non-reasoning)68.9
6Gemini 2.0 Flash68.6
7Qwen 3.5 4B (Non-reasoning)68.2
8Claude Haiku 4.566.8
9Qwen 3 VL 8B Instruct66.5
10Gemini 2.0 Flash Lite66.5
11GPT-5 Mini (Minimal)66.4
12Qwen 3 VL 4B Instruct66.1
13Nova 2 Lite66.1
14Gemini 2.5 Flash Lite64.6
15Molmo2-8B64.6

Interactive version: theaggregate.ai/benchmark?slug=prism-vlm-quality · How It Works · Data refreshed daily, snapshot 2026-09-26.