JoyCodevsKiro
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
JoyCode
JD.com · AI-native IDE
#74MCP
Panel
6.02 spec wins
- Reliability
- 6.0
- Usefulness
- 6.2
- Cost
- 5.7
- Longevity
- 6.2
“A cloud console shows your running remotes, apps and middleware, in case the one thing missing from your editor was an on-call dashboard.”
Kiro
Amazon Web Services · AI-native IDE
#57MCP
Panel
6.55 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.”
Spec by spec
| Spec | JoyCode | Kiro |
|---|---|---|
| Architecture | ||
| Category | AI-native IDE | AI-native IDE |
| Runs | local, cloud | local, cloud, sandbox |
| Platforms | macos, windows | macos, linux, windows, web |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | Yes | Yes |
| MCP server | No | No |
| Capabilities | ||
| Runs terminal commands | NoAgents are configured with built-in tools and MCP servers, but the published docs do not name a shell or terminal tool. | Yes |
| Multi-file edits | YesImplied by the agent and built-in tool model and the spec-driven end-to-end claim; the three published doc pages do not state it in those words. | Yes |
| Git operations | No | Yes |
| Browser control | No | 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. |
| Sandboxed execution | No | YesKiro Web runs each task in its own isolated execution environment with configurable access controls. |
| Multi-agent orchestration | YesA team defines multiple named agents with their own prompts, tools and MCP servers. | Yes |
| Headless / CI mode | No | Yes |
| Models | ||
| Backbone | bundled Chinese models, custom client-side models | Claude, GPT, DeepSeek, Qwen, MiniMax, GLM |
| Bring your own model | Yes | 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. |
| Local models | No | NoNo local or OpenAI-compatible endpoint is documented; only Kiro-hosted models are selectable. |
| Cost | ||
| Pricing model | seat | mixed |
| Starts at | n/a | $20/mo |
| Free tier | Yes | Yes |
| Bring your own key | Yes | NoKIRO_API_KEY for headless runs is a Kiro platform token billed against Kiro credits, not your own provider or AWS account key. |
| Openness | ||
| Open source | No | No |
| License | proprietary | proprietary |
| GitHub stars | n/a | 4,343 |
Which one would each critic pick
| Critic | JoyCode | Kiro | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 6.3 | 7.3 | Kiro — Pick Kiro if you want an agent that makes you write requirements and a design before it touches code, and accept that the bill is a credit meter you cannot route around with your own key. |
| El Crítico | 5.5 | 6.3 | Kiro — Kiro meters every task in credits with per-model multipliers and no key of your own, and the headless CLI ships a --trust-all-tools flag that turns off the only approval gate it has. |
| El Profesor | 6.5 | 6.5 | no preference |
| La Inversora | 6.3 | 7.8 | Kiro — Kiro is AWS's entry, and the question is never whether AWS can fund it, only whether AWS keeps caring. |
| La Jefa | 6.0 | 7.0 | 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 | 5.5 | 4.3 | JoyCode — Closed source, but the server configuration is a real mcpServers JSON block with documented npx and uvx invocations per operating system, and custom models attach on the client. |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.