# Microsoft's ParetoSaddler applies cost preference only at the end of agent harness search

> Original title: Super interesting finding on optimizing agent harnesses.

- Source: Elvis Saravia
- Published: 2026-11-03T21:58:01.000Z
- Bad Influence: https://badinfluence.news/items/m8ol7s28w0anthy85jickiclh
- Original: https://x.com/omarsar0/status/2117738094069944347

## Summary

Elvis Saravia summarizes Microsoft's ParetoSaddler paper, which optimizes agent harnesses for quality and cost. The method applies cost preference only at final deployment, not at every step, so it keeps candidates that later lead to better harnesses. On GAIA2, quality rose from 56.22% to 62.89% while cost per task fell from $0.78 to $0.45.

## Post

Super interesting finding on optimizing agent harnesses.

When you optimize an agent harness for both quality and cost, apply your cost preference only at the end of the search, since applying it at every step throws away harnesses that lead to better ones.

That is the key insight behind Microsoft's ParetoSaddler.

At each step, an optimizer edits a parent harness. Both versions run several times, and a Bayesian bootstrap over those runs decides whether the candidate shows credible progress in quality or in cost. Promoted candidates stay available as parents, and the final deployment choice is made separately.

On GAIA2, quality goes from 56.22% to 62.89% while cost per task drops from $0.78 to $0.45. On ClawMark, quality goes from 70.86% to 81.35%, and cost drops from $1.17 to $1.01.

The Pareto-selected harnesses also beat quality-only optimization on both quality and cost.

If your harness search only tracks accuracy, add repeated runs and a cost axis before deciding which candidates to keep.

https://arxiv.org/abs/2611.02079

Chat with Paper: https://academy.dair.ai/papers/paretosaddler-pareto-guided-optimization-of-agent-harnesses-for-quality-and-cost-2611.02079

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWO3kq0a4AEZYsA.png&mode=thumb&exp=1794009600&sig=26f3ade452c67304)