Microsoft Sico-Evolution 4B Model Achieves 82.9% GUI Success
June 30, 2026
Microsoft's Sico-Evolution 4B model achieves an 82.9% task success rate on GUI agents, significantly outperforming larger models like GPT-5.4 and Claude Opus 4.6.
HOW THIS AFFECTS YOU
●
builderYou can deploy highly capable GUI agents using significantly smaller, lower-latency 4B parameter models.
●
researcherSmall-scale models can achieve state-of-the-art performance in specialized agentic tasks through optimized training.