---
title: 'In the News: September 17, 2026'
description: 'A UC Berkeley benchmark finds coding-agent harness choice barely changes success rates but can double the bill, weakening the case for a model default harness.'
canonical_url: 'https://darkfactory.dev/news/2026-09-17-midday'
markdown_url: 'https://darkfactory.dev/news/2026-09-17-midday.md'
collection: news
date_published: '2026-09-17T13:20:00-04:00'
date_modified: '2026-09-17T13:20:00-04:00'
---

# In the News: September 17, 2026


UC Berkeley's Sky Lab tested 21 model-harness pairs and found that harness choice changed cost far more than success rates. Claude Code cost about twice as much as Pi across shared models on SWE-bench Lite, while another harness achieved the best observed success rate in nine of twelve comparisons.

## 1. Harness choice moves cost more than success rates

**[HarnessTax: How Much Does the Harness Matter for Coding Agents?](https://harnesstax.github.io/)** · Melissa Z. Pan, Shuo Yang, Negar Arabzadeh, Wei-Lin Chiang, Ion Stoica, Matei Zaharia, UC Berkeley and Arena · September 2026

The researchers tested seven models across three harnesses, Claude Code, Codex CLI, and the minimal open-source harness Pi, on SWE-bench Lite and Terminal-Bench 2.0: 21 pairs, 30 sampled tasks repeated three times each. Claude Fable 5 solved 97.8 percent of attempts in Claude Code versus 96.7 percent in both Codex and Pi, a gap within noise, yet Claude Code cost about twice as much per attempt: $1.33 against $0.67 in Pi. Across shared models, Claude Code cost roughly 2.0 times as much as Pi and 1.6 times as much as Codex on SWE-bench Lite, while the average effect of harness choice on success rate stayed within about 2 percentage points on that benchmark and about 5 points on Terminal-Bench 2.0. Some of the gap shows up before the agent does anything: across all seven models, Claude Code's mean initial context ran more than ten times Pi's, in instructions and tool schemas alone. The authors put it plainly: "you may be paying such a hidden 'harness tax' when you accept a coding agent's default harness without comparing alternatives." Across six Anthropic and OpenAI models on both benchmarks, a different harness produced the best observed success rate in nine of twelve comparisons, including Sonnet 4.6 solving more tasks in Codex than in Claude Code at a similar cost. The authors say they will publicly release their profiling traces.

**Why it matters:** Teams standardizing a factory's coding agent should benchmark the model and harness separately. In this study, harness choice affected cost much more than task success, and vendor-matched pairings did not reliably lead the comparisons.