Kiro
Overview
Kiro is an agentic AI IDE that turns plain prompts into structured, testable work. It transforms natural language intent into EARS‑style requirements and acceptance criteria, proposes architecture and system design grounded in best practices, and breaks delivery into dependency‑aware tasks with optional tests. Agents run where developers work: in the terminal, locally or over SSH, executing commands, reading outputs, and applying fixes with diff review. Native MCP support connects docs, databases and APIs (including remote), while steering files encode project rules and context to guide behavior. With autopilot and event‑driven hooks, Kiro can run long jobs and background automation safely, and per‑prompt credit visibility keeps costs predictable. Built on Code OSS, it stays compatible with Open VSX plugins and familiar settings so teams can adopt it with minimal change.
Use Cases
Requirements from Prompts
Turn natural prompts into EARS‑style requirements and acceptance criteria.
Architecture & Design
Generate architecture and system design aligned with your codebase and best practices.
Terminal Agent Workflows
Execute commands, trace errors, and ship features from the terminal.
Agent Hooks Automation
Trigger background agents on events to generate docs, tests, and optimizations.
Multimodal Context
Use images and files to steer implementation from UI mockups to whiteboards.
Credit Tracking
See per‑prompt credit usage to stay within budgets.
Pricing
Individual Plans
Includes:
- 50 credits
Includes:
- 1,000 credits
- Pay‑per‑use overage ($0.04/credit)
Includes:
- 2,000 credits
- Pay‑per‑use overage ($0.04/credit)
Includes:
- 10,000 credits
- Pay‑per‑use overage ($0.04/credit)
Business Plans
Includes:
- Assign paid plans to individual users
- Centralized team billing
- Usage analytics and reporting
- SAML/SCIM SSO via AWS IAM Identity Center
- Pay‑per‑use overage
- Organizational management dashboard
- Enterprise security and privacy controls
Core Features
Standard Features
Advanced Features
Executable Specs
Drive agent workflows from structured, executable specs that make intent explicit.
EARS Requirements
Turn intent into precise EARS requirements and acceptance criteria.
Dependency‑Aware Plans
Sequence tasks by dependencies with optional tests for safer delivery.
Native Terminal Agents
Run agents in the terminal locally or over SSH with interactive loops.
Autopilot Mode
Let agents run long jobs safely with approval gates when needed.
Agent Hooks
Trigger background automation on events like file save to scale routine work.
MCP Integration
Connect docs, databases and APIs (including remote) to enrich capabilities.
Diff‑First Review
See code diffs as they happen and approve or edit with one click.
Multimodal Context
Use images and files to steer implementation from mockups to whiteboards.
Credit Visibility
Track per‑prompt credit usage in real time to manage costs.
AI Models Support
Built-in Models
Anthropic
API Integration
Custom API
System Requirements
Supported Platforms
Windows, macOS, Linux
API Support
MCP servers and terminal integrations
Compatibility
Built on Code OSS; works with Open VSX plugins
Similar Products
Loading similar products...