Kiro
AWS's agentic, spec-driven AI IDE that turns prompts into requirements, designs, and sequenced tasks, then builds them with parallel agents
Kiro is AWS's agentic, spec-driven AI IDE built on Code OSS, generally available since May 2026. It turns prompts into requirements, designs, and sequenced tasks, then implements them with parallel agents. Billing is credit-based: Free includes 50 credits, Pro is $20/month for 1,000, scaling to Power at $200/month for 10,000, with $0.04 overage per credit. Best for structured, spec-first delivery.
What is Kiro?
Kiro is AWS’s agentic AI IDE, built on Code OSS and generally available since May 2026 after an earlier preview and waitlist period. Its defining idea is spec-driven development: rather than generating code straight from a prompt, Kiro first turns your request into a spec — a set of formal requirements, an architectural design, and a sequenced list of tasks. You review and adjust that plan, and only then do agents write the code. The spec remains as living documentation, which makes the approach well suited to larger or longer-lived projects where traceability matters more than raw speed.
Once a spec is approved, Kiro runs parallel agents that implement multiple tasks concurrently across a codebase, backed by property-based testing that asserts rules which must hold across all inputs rather than checking a few hand-written cases. Hooks let you trigger background automations on events like file save, while steering files carry shared project context across Kiro’s IDE, CLI, and web interfaces. Because it is built on Code OSS, you can import VS Code settings, themes, and Open VSX plugins, and it supports MCP for external tools. Billing is credit-based with no daily or weekly rate limits: an agentic interaction consumes credits, inline autocomplete does not, and overage is capped at $0.04 per credit so spend stays predictable. Multiple models are available, including Claude Sonnet and OpenAI options.
Who is it for?
Kiro fits developers and teams who prefer structure over improvisation — people who want an AI agent to plan work into reviewable requirements and tasks before it starts editing files. The Free tier’s 50 credits are enough to evaluate the spec workflow, but the real value shows up on paid plans running parallel agents against real codebases. If you want a lighter, chat-first agentic editor, Cursor is closer to that model, and for pure agentic coding many developers still reach for Claude directly.
- Teams building larger, longer-lived projects who want requirements and design docs generated as a byproduct of implementation, improving traceability and onboarding.
- AWS-aligned engineering orgs that value enterprise-grade security, SSO, and compliance controls from the vendor operating the tool.
- Developers migrating from VS Code who want to keep their settings, themes, and Open VSX plugins while adopting agentic, spec-first workflows.
- Practitioners who prefer planning before code — writing and reviewing a spec first, then delegating sequenced tasks to parallel agents.
How much does Kiro cost?
Starting price: $20/mo · Free tier: yes · Model: freemium
Price history tracked from June 2026
| Plan | Price | Includes |
|---|---|---|
| Free | Free | 50 monthly credits · Open-weight models plus Claude Sonnet 4.5 · Specs, hooks, and steering files · No credit card required |
| Pro | $20/mo | 1,000 monthly credits · Premium model access · Add-on credits at $0.04 each · Overage capping controls |
| Pro+ | $40/mo | 2,000 monthly credits · Premium model access · Add-on credits at $0.04 each |
| Pro Max | $100/mo | 5,000 monthly credits · Premium model access · For heavy agentic workloads |
| Power | $200/mo | 10,000 monthly credits · Premium model access · Highest credit allocation |
| Teams | $20/user/mo | Same credit tiers as individual plans · Consolidated billing and usage analytics · SSO and organizational management |
What are Kiro's key features?
- Specs — converts prompts into requirements, architectural designs, and sequenced tasks
- Parallel agents that execute implementation tasks concurrently across large codebases
- Hooks and automations that run background tasks on predefined triggers
- Steering files that carry context across the IDE, CLI, and web interfaces
- Property-based testing that asserts rules which must hold across all inputs
- MCP (Model Context Protocol) support for external tool integration
- Built on Code OSS with VS Code settings, themes, and Open VSX plugin compatibility
- Multi-model access including Claude Sonnet and OpenAI models
What people use Kiro for
- 01 Turning a rough product prompt into formal requirements, an architecture design, and sequenced implementation tasks before any code is written
- 02 Running parallel agents to implement multiple tasks concurrently across a large codebase
- 03 Automating repetitive maintenance with hooks that trigger background tasks on events like file save
- 04 Carrying shared project context across the IDE, CLI, and web through steering files
- 05 Migrating from VS Code by importing existing settings, themes, and Open VSX plugins
Pros and cons
| Pros | Cons |
|---|---|
| Spec-first workflow produces requirements and design docs as a byproduct, improving traceability on larger projects | Credits are consumed by every agentic interaction and do not roll over; heavy multi-step tasks can exhaust a monthly allocation and push you into $0.04-per-credit overage |
| Built on Code OSS — VS Code settings, themes, and Open VSX plugins import directly | The spec-driven workflow adds upfront overhead — writing and reviewing requirements and design docs before coding is slower for small, throwaway tasks |
| Free tier includes 50 credits and Claude Sonnet 4.5 access with no credit card | Relatively new (GA in May 2026), so plugin coverage, community, and ecosystem are less mature than established editors |
| Backed by AWS with enterprise-grade security, SSO, and compliance controls | AWS-operated, so some teams will weigh vendor lock-in and account requirements |
| Credit billing has no daily or weekly rate limits, with overage capping to control spend |
What are the best Kiro alternatives?
Trae is the closest Kiro alternative in our directory: it covers the same category (AI Coding), starts at $3/mo, has a free tier.
Ranked by category overlap with Kiro, then free-tier availability, then lowest verified starting price — computed from our verified data, never from sponsorships.
| Alternative | What it is | Starting price | Free tier | Price verified |
|---|---|---|---|---|
| Trae | ByteDance's free, VS Code-based AI code editor with an autonomous SOLO mode that scaffolds and builds full-stack projects from a prompt | $3/mo | yes | Jun 30, 2026 |
| GitHub Copilot | AI coding assistant with inline completions, chat, and autonomous agent mode across major IDEs | $10/mo | yes | Jul 3, 2026 |
| Windsurf | Agentic AI IDE with Cascade that writes, edits and runs code autonomously inside your editor | $15/mo | yes | Jun 11, 2026 |
| Claude | Anthropic's AI assistant built for long-context analysis, coding, and safe reasoning | $20/mo | yes | Jul 3, 2026 |
| Cursor vs Kiro → | AI-native code editor that runs autonomous agents to build, test, and review code | $20/mo | yes | Jul 3, 2026 |
Head-to-head comparisons
How people make money with Kiro
- Spec-driven delivery retainer — offer clients a structured requirements-to-code service where Kiro's specs double as living project documentation
- Freelance agentic development gigs, using parallel agents to ship features faster while billing at standard project rates
Frequently asked questions
Is Kiro free?
Yes. The Free tier costs $0 and includes 50 monthly credits, access to open-weight models plus Claude Sonnet 4.5, and core features like specs, hooks, and steering files. No credit card is required to start.
How much does Kiro Pro cost?
Kiro Pro is $20/month per user and includes 1,000 monthly credits with premium model access. If you run out, add-on credits are billed at $0.04 each, and you can set caps to control overage spend.
What are Kiro's paid tiers?
Beyond Free, Kiro offers Pro at $20/month (1,000 credits), Pro+ at $40/month (2,000 credits), Pro Max at $100/month (5,000 credits), and Power at $200/month (10,000 credits). All paid tiers share the $0.04-per-credit overage rate.
Is Kiro generally available?
Yes. Kiro moved past its 2025 preview and waitlist to general availability in May 2026. It is a shipping product, not a beta, and continues to add models such as Claude Sonnet and OpenAI options.
What is spec-driven development in Kiro?
Instead of jumping straight to code, Kiro first converts your prompt into a spec: formal requirements, an architectural design, and a sequenced task list. You review that plan, then agents implement the tasks. The spec stays as living documentation.
Do Kiro credits roll over?
No. Monthly credits reset each billing cycle and do not roll over. Purchased add-on credits are valid for 12 months. An agentic interaction — generating a spec, running a hook, or executing a multi-step task — consumes credits, while inline autocomplete does not.
Is Kiro built on VS Code?
Kiro is built on Code OSS, the open-source base of VS Code. You can import your existing VS Code settings, themes, and Open VSX-compatible plugins, so migrating your editor setup is straightforward.
Does Kiro work with Claude?
Yes. Kiro provides multi-model access, and Claude Sonnet is available even on the Free tier, with premium models on paid plans. Developers who want Claude's agentic coding directly can also use Claude on its own.