CodeArts AgentvsZed
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
CodeArts Agent
Huawei Cloud · AI-native IDE
#44MCP
Panel
6.81 spec wins
- Reliability
- 7.0
- Usefulness
- 6.8
- Cost
- 5.8
- Longevity
- 7.5
“There is a HarmonyOS agent and an ArkTS-tuned model, available on the China site only, which is a very specific way to be excluded.”
Zed
Zed Industries · AI-native IDE
#2OSSMCP
Panel
7.54 spec wins
- Reliability
- 7.2
- Usefulness
- 6.7
- Cost
- 8.0
- Longevity
- 8.0
“A GPL editor that will host Claude, Codex and Gemini in one panel, so the agent wars now happen inside a Rust window.”
Spec by spec
| Spec | CodeArts Agent | Zed |
|---|---|---|
| Architecture | ||
| Category | AI-native IDE | AI-native IDE |
| Runs | local, cloud | local |
| Platforms | macos, linux, windows | macos, linux, windows |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | Yes | Yes |
| MCP server | Yes | No |
| Capabilities | ||
| Runs terminal commands | YesThe v26.5.1 release notes describe a CLI in which you run commands and a TUI in which you use natural language to complete coding tasks. | Yes |
| Multi-file edits | Yes | Yes |
| Git operations | YesGit operations are listed among the coding tasks the CLI completes, and the IDE reviews Git commit diffs. | Yes |
| Browser control | No | NoThe agent's only web tools are fetch, which returns a URL as Markdown, and search_web; there is no browser control, Playwright or DevTools protocol access. |
| Sandboxed execution | No | YesAgent tool calls run in an OS-level sandbox — Seatbelt via sandbox-exec on macOS and bubblewrap on Linux, applied to the terminal and fetch tools; on Windows only inside WSL. |
| Multi-agent orchestration | YesSub-agents are created and executed automatically for task decomposition and delegation, and multiple sessions run concurrently. | Yes |
| Headless / CI mode | No | NoThe zed CLI only opens files, windows and diffs — no print, exec or other non-interactive mode for running the agent from a script or CI is documented. |
| Models | ||
| Backbone | GLM-5.2, openPangu-2.0-Flash, openPangu-2.0-Pro, GLM-5.2-ArkTS-SPARK | Claude, GPT, Gemini, Bedrock, OpenRouter, Ollama, LM Studio, Copilot |
| Bring your own model | YesThird-party models can be configured by the individual on the entry and basic tiers, and centrally by an administrator on the professional tier. | Yes |
| Local models | No | Yes |
| Cost | ||
| Pricing model | mixed | mixed |
| Starts at | n/a | $10/mo |
| Free tier | Yes | Yes |
| Bring your own key | No | Yes |
| Openness | ||
| Open source | No | Yes |
| License | proprietary | GPL-3.0-or-later |
| GitHub stars | n/a | 91,181 |
Which one would each critic pick
| Critic | CodeArts Agent | Zed | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 7.0 | 7.8 | Zed — Pick Zed if you live in the editor and want the agent to feel instant; pick Cursor if you want the agent to do more of the work for you. |
| El Crítico | 6.3 | 7.0 | Zed — The agent runs a terminal tool on your working tree with sandboxing off by default, and the fetch tool stays outside the sandbox even when you turn it on. |
| El Profesor | 7.0 | 7.3 | Zed — No benchmark claim, and an architecture worth studying: the Agent Client Protocol lets foreign agents run inside the editor, so the design does not depend on Zed's own loop. |
| La Inversora | 7.5 | 7.3 | CodeArts Agent — Personal tiers run CNY 0, 98, 198 and 498 a month, and the on-demand meter is denominated in model tokens. The plugin is a funnel into the parent's inference business. |
| La Jefa | 7.5 | 6.5 | CodeArts Agent — Single sign-on, IP allow-lists, per-scope governance and token quotas, with a seat-based professional edition sized from 1 to 1000. This is the first row this quarter written for me. |
| El Hacker | 5.5 | 9.0 | Zed — GPL-3.0, MCP servers in a context_servers block in settings.json, Ollama and LM Studio as providers, and ACP to plug my own agent in. |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.