← Back

Wire · opportunities

Einsia AI Launches Visual Cipher Bench to Evaluate Long-Horizon Visual Reasoning in AI Agents

Published

18 November 2026

Topic

opportunities

◆ Sectors

AI & ML

◆ Geography

Singapore

◆ Source

Read at morningstar.com →

◆ Verified

Fusion42 · 18 November 2026 · Fusion42 review

Einsia AI has launched Visual Cipher Bench, a new benchmark to evaluate long-horizon visual reasoning in AI agents, focusing on autonomous rule discovery, logical planning, and spatial operations. Testing eight advanced AI models, the benchmark revealed significant deficiencies, with the best model achieving only 26.7% accuracy, highlighting the gap between executing known instructions and independently discovering hidden rules within complex visual puzzles.

This Wire brief sits within Fusion42's coverage of AI & ML, and 3 sources have reported it.

◆ ◆ The Wire takeaway

You can no longer rely on AI models to autonomously discover complex rules in visual tasks; your AI product must explicitly incorporate rule guidance or risk poor performance. This gap opens a new avenue to build tools that help AI agents validate and apply hidden instructions effectively.

◆ Coverage

3 sources · 18 Nov 2026

◆ Related on Wire

◆ Topics

AI & MLai-agentsvisual-reasoningrule-discoverybenchmarkgpt-6