Agensh: Scaling Organizational Intelligence to 1,024 Agents
Agensh lets up to 1,024 self-organized agents raise ProgramBench scores without a central orchestrator.
Agensh is a self-organized multi-agent harness that avoids a central orchestrator. Workers concurrently gather context, claim sub-tasks, act, share findings, verify results, and merge progress through a shared workspace, message interface, and shared context. On the five hardest ProgramBench tasks with GPT-5.6-sol (high), scaling from 1 to 128 agents raised the mean final test-pass rate from 19.31% to 28.78%. On pandoc, scaling from 1 to 1,024 agents raised the final test-pass rate from 33.89% to 55.06%.
- Agensh has no central orchestrator; workers claim and self-assign tasks asynchronously.
- Shared workspace, messaging, and shared context support the cooperation loop.
- With 128 agents, mean ProgramBench pass rate rose from 19.31% to 28.78%.
- On pandoc, 1,024 agents raised the pass rate from 33.89% to 55.06%.
- The evaluation used GPT-5.6-sol at the high setting.
Full article238 words · extracted from arxiv.org · click to collapse
A multi-agent system can reduce latency on complex tasks by executing work concurrently. Several pioneering harness frameworks support multi-agent systems. However, the scalability of current multi-agent harnesses is often constrained by a central orchestrator's capacity to allocate tasks and coordinate workers. To address this limitation, we introduce Agensh, a scalable self-organized multi-agent harness without a central orchestrator: concurrent workers execute a multi-agent cooperation loop, continuously gathering context, claiming and self-assigning sub-tasks, taking action and sharing findings, verifying results, and merging progress in an asynchronous manner. The loop is supported by the agentic organization infrastructure comprising three components: a shared workspace holds proposed, ongoing, and completed work; a message interface lets workers communicate; and shared context retains reusable findings and work intentions. To test the scalability of Agensh, we evaluate it on the five hardest ProgramBench tasks with GPT-5.6-sol (high). Scaling from 1 to 128 agents raises the mean final test-pass rate from 19.31% to 28.78%, an approximately 49% relative improvement. Larger organizations reach comparable test-pass rates earlier. On pandoc, scaling from 1 to 1,024 agents raises the final test-pass rate from 33.89% to 55.06%. Worker trajectories further show that different forms of self-organized cooperation gradually emerges and standardizes as the organization grows. These results reveal the number of agents as a new scaling dimension for multi-agent organizations to expand the frontier of general intelligence, offering a practical solution for complex tasks under hard latency constraints or time budgets.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arxiv.org/abs/2609.26781