Harness Active Open source
ZCode
Z.ai's agentic development environment, tuned for GLM models.
Get ZCode Install
Download ↗
opens in a new tab
Pricing
The app is free to download and first-time users get a 5-day trial with 5M tokens a day. Continued use requires a GLM Coding Plan or your own API key; idle-time tasks run without consuming plan quota.
- Lite $18/mo 10,000 credits/week; $12.6/mo billed yearly
- Pro $80/mo 6x Lite usage; $56/mo billed yearly
- Max $168/mo 14x Lite usage; $117.6/mo billed yearly
- Z.ai API Pay as you go Pay per token through the Z.ai open platform See plan ↗
Prices as listed on 19 Oct 2026.
AboutZ.ai (Zhipu AI)
ZCode is Z.ai's Agentic Development Environment built around its first-party ZCode Agent and tuned for the GLM-5.3 model family. It ships as a desktop app with a built-in browser panel, subagents, a plan mode, scheduled automations, hooks, skills and MCP, and can be controlled from a phone through Remote Control. The client is open source under Apache-2.0.
Features Verified 12/20 ✓ yes ✗ no ? not verified What do these mean?
Agents
3/5 verified ✓
Multi-provider
Runs agents from several providers side by side?
yes
?
ACP
Speaks the Agent Client Protocol?
not verified
?
Parallel agents
Runs several agent sessions at the same time?
not verified
✓
Subagents
Can an agent delegate work to child agents?
yes
✓
Custom agents
Can you define your own agent roles?
yes
Workflow
3/6 verified ✓
Plan mode
Has an explicit plan / approve-before-edit step?
yes
?
Worktrees
Keeps parallel work in separate git worktrees?
not verified
✓
Checkpoints
Can you rewind code and chat to an earlier turn?
yes
?
Diff review
Includes a built-in diff or review screen?
not verified
?
PR integration
Opens and tracks pull requests from the session?
not verified
✓
Automations
Can it start agent runs on a schedule or event?
yes
Integrations
4/5 verified ✓
MCP
Connects to MCP servers and their tools?
yes
✓
Skills
Can you add reusable skills or commands?
yes
✓
Hooks
Can your own scripts run on agent events?
yes
?
Issue tracker
Can it pull work from GitHub or Linear issues?
not verified
✓
Browser preview
Can the agent open and drive a web browser?
yes
Access & safety
2/4 verified ✓
Remote access
Can you follow or control sessions from another device?
yes
?
Cloud sandbox
Runs work in a hosted sandbox rather than on your machine?
not verified
✓
Local sandbox
Does it sandbox what agents can do locally?
yes
?
Headless / API
Can you run it from scripts, CI, or your own code?
not verified
Other harnessesSee all
