中Start a project
Released · Open-source Beta

ZyHive

Turn AI conversations
into an organized team.

Manage AI members, models, tools, memory, tasks, and messaging channels in one place. Source code, releases, installation, updates, and current operating boundaries are publicly available.

Install ZyHiveView source
ZyHive Workspace
Z1Plan goal
Z2Use tools
Z3Deliver result

Product architecture

A shared workspace for AI members, models, tools, and tasks.

ZyHive connects product design, model and tool integration, permission boundaries, CI, cross-platform release, installation, updates, and public operating limits.

01

Members & Organization

Give each AI member an identity, workspace, and responsibility to create a clear collaboration structure.

02

Memory & Context

Maintain context for long-running work through layered memory and project workspaces.

03

Tools & Permissions

Manage tools, approvals, and operating boundaries in one place so automated actions remain controlled.

04

Tasks & Goals

Keep configured work moving through task dispatch, schedules, and goal management.

05

Messaging Channels

Connect channels such as Telegram and Lark so teams can collaborate through existing work entry points.

06

Usage & Cost

Track model tokens and cost to understand the operating economics of an AI team.

Install

One command to start self-hosting.

Officially supports macOS and Linux on x86_64 and ARM64. The current product is a single-machine, self-hosted Beta.

TerminalStable: checking
$ curl -fsSL https://install.zyling.ai/install | bash

Current boundary

A public Beta with public boundaries.

ZyHive is suitable for individuals and small teams testing in local, private-network, and controlled environments. Models, cloud services, and third-party channels may incur separate costs.

  • Official Windows support is not currently promised
  • Multi-tenancy and multi-instance management are not currently promised
  • Enterprise-grade high availability is not currently promised
  • Complete an independent security and continuity assessment before critical use
  • The open-source edition is licensed under AGPL-3.0

Building an enterprise AI workflow with ZyHive?

A controlled pilot can define the required models, tools, permissions, and acceptance criteria.

Start a project