@rohanpaul_ai: So much recent work and research papers points to the same thing: the "harness" is becoming the real capability layer. …
Summary
Offloop's multi-agent harness achieves state-of-the-art results on GDPval, GDP.pdf, and JobBench benchmarks, outperforming Claude Code and Codex at 1/3 to 1/10 the cost per task, targeting $2.4T in US knowledge work.
View Cached Full Text
Cached at: 07/24/26, 09:05 AM
So much recent work and research papers points to the same thing: the “harness” is becoming the real capability layer.
@Offloop ’s 4-person team demonstrated a multi-agent harness outperforming Claude Code and Codex on GDPval benchmarks, targeting $2.4T in US knowledge work.
- The team scored 84.9 at $1.65 per task.
- Opus 4.8 inside Claude Code scored 82.4, and GPT 5.6 Sol inside Codex scored 83.3, costing far more per task, $14.38 and $5.20
GDPval measures how well AI handles real work across 44 occupations and 9 major industries. Models get shell access and web browsing, then face blind pairwise comparisons against human experts.
And those tasks map onto US jobs paying roughly $2.4T a year.
Offloop (@Offloop): In our internal evaluations, Offloop achieved new state of the art results on three benchmarks evaluating professional knowledge work:
84.9 on GDPval, 44 on GDP.pdf, 67.2 on JobBench - each ahead of Codex and Claude Code, at 1/3 to 1/10 of their cost per task.
Lower cost per
Similar Articles
@dair_ai: // State-Externalizing Harnesses // A new paradigm is emerging on how to effectively build agents and harnesses. If the…
Harness-1 introduces a state-externalizing harness that separates routine bookkeeping from policy decisions in search agents, enabling a 20B model to outperform larger frontier searchers across multiple benchmarks.
@AlphaSignalAI: https://x.com/AlphaSignalAI/status/2074130508833845396
Self-improving harnesses enable AI agents to autonomously rewrite their operating rules by analyzing execution traces, leading to a 60% performance boost. Research from Shanghai AI Lab introduces the Self-Harness framework, allowing lightweight models to outperform larger ones without manual engineering.
@omarsar0: // Self-Harness: Harnesses That Improve Themselves // (bookmark this one) Most of the agent scaffolds we rely on today …
This paper introduces Self-Harness, a new paradigm where LLM-based agents iteratively improve their own operating harness—prompts, tools, and control flow—without human engineers or stronger external agents, achieving significant performance gains across multiple models.
Own the Loop: A Field Guide to Agent Harnesses (5 minute read)
As AI coding models become commoditized, the agent harness—the control loop managing tools and workflows—emerges as the key differentiator. This guide maps the field of harnesses, weighing vendor-native performance against the portability of model-agnostic workflows.
@rohanpaul_ai: This paper shows that agent performance depends less on prompts alone and more on the harness around them. “Agent intel…
This paper argues that AI agent performance depends more on the harness (control layer) than on prompts alone, proposing natural-language agent harnesses to make design choices inspectable and portable.