According to Gartner research, open-source language models can handle approximately 90 percent of agentic work tasks while costing roughly one-twentieth as much as frontier large language models. The analysis suggests organizations may achieve significant cost efficiency by deploying open models for agent-based applications rather than relying exclusively on premium commercial alternatives.