Now with 5 AI providers — Anthropic, OpenAI, Google, xAI, Mistral

Build full-stack apps
with your words.

Start instantly with included credits — or connect your own AI keys and pay API rates directly. Either way, you own your code and your costs.

See how it works ↓

No credit card required · Export anytime · Cancel any time

app/page.tsx
import { Button } from "@/components/ui/button"

export default function Home() {
  return (
    <main className="flex min-h-screen items-center">
      <h1 className="text-4xl font-bold">Hello, World!</h1>
      <Button>Get Started</Button>
    </main>
  )
}
Generating…TypeScript · UTF-8

Sound familiar?

Paying for broken code

Credit-based tools charge you whether the generation works or not. A failed prompt still costs you.

No idea what changed

Most tools apply AI changes silently. You only find out something broke after the fact.

AI forgets your project

Hit the context limit and the AI loses track of everything it built. You start explaining from scratch.

We built this because we felt this too.

Three steps from prompt to working code.

1

Choose your stack

Pick your AI provider, database, auth, and deployment. Your keys. Your infrastructure.

2

Describe what you want

Type a prompt. Watch code stream into a real Monaco editor. Review the diff before it applies.

3

Deploy live

One click to Vercel, Railway, Render, or Netlify. Or share a preview link instantly.

See it in action.

Watch a full app go from prompt to working preview.

Demo video coming soon

We're recording a full walkthrough — prompt to working preview in one take. Join the waitlist and we'll send you the link when it's live.

Everything you need. Nothing you don't.

Bring Your Own API Key

Connect Anthropic, OpenAI, Google, xAI, or Mistral. Pay API rates directly — never our markup.

Diff View Before Apply

See exactly what the AI wants to change before it touches your code. Approve or discard.

Full Code Ownership

Export your entire project as a zip anytime. No hostage code. No lock-in. It's yours.

Smart + Flash Models

Routine edits auto-route to fast cheap models. Complex features use the best model available.

Share Preview Links

Get a public URL for your sandbox in one click. Share before you deploy. 2-hour live link.

Your Stack, Your Rules

Choose your own database, auth, and deploy provider. We generate code that fits your exact setup.

How we compare.

An honest look. All competitors are great tools — here's where we're different.

FeatureGreybeardAIReplitLovableBolt.new
Typical gen costtypical~$0.04–$0.40~$0.30–$1.00+~$0.30–$0.60+~$0.40–$0.80+
Pricing modelPay API costCreditsCreditsCredits
Failed gen cost~$0 extraFull creditFull creditFull credit
Cost per generation shownLive ✓
Choose AI provider5 providers112
Diff view before applyYesPartial
Export projectFree ZIP alwaysYesGitHub onlyYes
Stack picker (DB / Auth / Deploy)YesPartial
Push to GitHubYesYes
Share live preview linkYesYes
One-click deployYesYesPartialPartial

Verified March 2026 · Cost estimates based on public pricing pages · "Partial" = available on higher plans or extra steps

Simple pricing. No surprises.

Your AI costs go directly to your provider. We only charge for the platform.

MonthlyAnnual

Indie

$29/mo

For solo builders

  • 5 projects
  • All 5 AI providers
  • 50 sandbox-hours/month
  • Project memory
  • ZIP export
Most Popular

Pro

$79/mo

For serious builders

  • Unlimited projects
  • 100 sandbox-hours/month
  • All providers + smart models
  • Priority generation queue
  • Cost analytics

Teams

$199/mo

For teams building together

  • 8 seats
  • Shared stack config
  • All Pro features
  • Team dashboard
  • Audit logs
  • SSO

All plans include a 14-day free trial · No credit card required · Cancel anytime

AI costs paid directly to your provider. Typical generation: ~$0.001 (Flash) to ~$0.02 (Smart).

Questions we get asked.

You only pay your AI provider's API cost — typically under $0.02. You never pay us for a failed generation. Our platform fee is a flat monthly subscription.

Yes — anytime, for any project. Click Export in the workspace to download a complete zip with all files, package.json, and a .env.example. No lock-in, ever.

Anthropic (Claude), OpenAI (GPT), Google (Gemini), xAI (Grok), and Mistral. Each with Smart and Flash tiers.

No. 14-day free trial, no card required. You'll need an API key from your chosen AI provider.

Next.js 14, TypeScript, and Tailwind CSS. You choose your own database, auth, and deploy provider.

Flash models (Haiku, GPT-4o-mini, Codestral) handle routine edits quickly and cheaply. Smart models handle complex new features. We pick automatically, or you can override per prompt.

Yes on Pro and Teams plans. Connect a custom domain in the deployment settings after deploying.

Fair question — and honestly, if you need something today, use them. They're good tools. We're building for a different priority: you bring your own API key and pay your AI provider directly, so a failed generation costs you fractions of a cent rather than burning credits. You see every file change as a diff before it's applied. And you choose your AI provider — Anthropic, OpenAI, Google, xAI, or Mistral — rather than being locked to one. If those tradeoffs matter to you, we think the wait is worth it. If they don't, Bolt.new is a solid choice right now.

Start building your first app today.

14-day free trial. No credit card. Real code you own.

Join the waitlist — invites rolling out weekly