KirovsZed
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
Kiro
Amazon Web Services · AI-native IDE
#57MCP
Panel
6.52 spec wins
- Reliability
- 6.3
- Usefulness
- 6.7
- Cost
- 5.7
- Longevity
- 7.3
“Makes you write requirements, design and tasks before any code, so the credit meter starts running before the first line does.”
Zed
Zed Industries · AI-native IDE
#2OSSMCP
Panel
7.56 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 | Kiro | Zed |
|---|---|---|
| Architecture | ||
| Category | AI-native IDE | AI-native IDE |
| Runs | local, cloud, sandbox | local |
| Platforms | macos, linux, windows, web | macos, linux, windows |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | Yes | Yes |
| MCP server | No | No |
| Capabilities | ||
| Runs terminal commands | Yes | Yes |
| Multi-file edits | Yes | Yes |
| Git operations | Yes | Yes |
| Browser control | YesEvery Kiro Web sandbox ships a headless Chrome with Playwright MCP, the agent-browser CLI and the Chrome DevTools MCP server preinstalled; the local IDE and CLI have no built-in browser tool. | 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 | YesKiro Web runs each task in its own isolated execution environment with configurable access controls. | 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 | Yes | Yes |
| Headless / CI mode | Yes | 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 | Claude, GPT, DeepSeek, Qwen, MiniMax, GLM | Claude, GPT, Gemini, Bedrock, OpenRouter, Ollama, LM Studio, Copilot |
| Bring your own model | NoKiro exposes a picker over its own curated model list plus an Auto router; no custom base URL, Bedrock deployment or self-hosted endpoint is documented. | Yes |
| Local models | NoNo local or OpenAI-compatible endpoint is documented; only Kiro-hosted models are selectable. | Yes |
| Cost | ||
| Pricing model | mixed | mixed |
| Starts at | $20/mo | $10/mo |
| Free tier | Yes | Yes |
| Bring your own key | NoKIRO_API_KEY for headless runs is a Kiro platform token billed against Kiro credits, not your own provider or AWS account key. | Yes |
| Openness | ||
| Open source | No | Yes |
| License | proprietary | GPL-3.0-or-later |
| GitHub stars | 4,343 | 91,181 |
Which one would each critic pick
| Critic | Kiro | Zed | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 7.3 | 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 | 6.5 | 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 | 7.8 | 7.3 | Kiro — Kiro is AWS's entry, and the question is never whether AWS can fund it, only whether AWS keeps caring. |
| La Jefa | 7.0 | 6.5 | Kiro — SSO and centralized billing exist, steering files make rules enforceable, and the meter has an admin-controlled overage switch; approved with conditions. |
| El Hacker | 4.3 | 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.