# Open-weight models could cut routine enterprise AI costs while safeguarding data

- URL: https://talkfuse.com/story/open-models-can-do-90-percent-of-agentic-work-at-a-twentieth-of-the-cost-of-a-frontier-llm-gartner/
- Published: 2026-11-03T06:19:50+00:00
- Updated: 2026-11-03T06:32:16+00:00
- Original publisher: The Register
- Original article: https://www.theregister.com/ai-and-ml/2026/11/03/open-models-can-do-90-percent-of-agentic-work-at-a-twentieth-of-the-cost-of-a-frontier-llm-gartner/5307132
- Topics: AI, Art, Culture, Technology

## Key points

- Open-weight models could handle nine in ten routine agent steps.
- Practical savings may be five or six times versus frontier models.
- Gartner expects open models to run 90% of these steps by 2030.

## Summary

Gartner analysts Mark Petty and Raymond Paquet argue that open-weight models could handle most routine steps in enterprise AI agent workflows, leaving costly frontier models for tasks that need more advanced reasoning. They say this matters both for reducing costs and for keeping sensitive workflow data within an organisation’s control.

The analysts estimate that nine in ten workflow steps are routine, including classification, tool use and data checks. At the same performance score, open models can cost a quarter as much as the cheapest closed model, or as little as a twentieth of using a frontier model for every step; allowing for extra retries, the practical saving may be five or six times. Gartner’s planning assumption is that open-weight models will run 90% of enterprise operational agent steps by 2030, with software harnesses that route and manage models becoming a key source of value.

---
Summarised by TalkFuse (https://talkfuse.com/). Quote the original publisher for the reporting and link this page for the summary and discussion.