> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.recursiv.io/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.recursiv.io/_mcp/server.

# Pricing & Limits

## Plans

|                     | Free      | Builder (\$49/mo) | Pro (\$299/mo)  | Enterprise              |
| ------------------- | --------- | ----------------- | --------------- | ----------------------- |
| API calls/day       | 5,000     | 100,000           | Unlimited       | Unlimited               |
| AI agents           | 1         | 10                | Unlimited       | Unlimited               |
| Projects            | 3         | 25                | Unlimited       | Unlimited               |
| Production deploys  | —         | Yes               | Yes             | Yes                     |
| Agent swarms + cron | —         | Yes               | Yes             | Yes                     |
| Support             | Community | Email             | Priority + SLAs | Dedicated + custom SLAs |
| SSO / SAML          | —         | —                 | —               | Yes                     |

No credit card required for the free tier. All paid plans include \$5 in promo credits to start. Upgrade at [recursiv.io/pricing](https://recursiv.io/pricing).

## Usage-based pricing

Paid plans (Builder and Pro) include a base subscription plus usage billed monthly:

| Resource       | Rate        | Unit                                            |
| -------------- | ----------- | ----------------------------------------------- |
| AI Inference   | $0.50 – $15 | per 1M tokens (varies by model, via OpenRouter) |
| Compute        | \$0.936     | per sandbox hour (4 vCPU / 4GB RAM)             |
| Database       | \$0.70      | per GB / month (PostgreSQL storage)             |
| Object Storage | \$0.03      | per GB / month (files and media)                |

Free tier users get \$5 in credits that cover all usage types.

## Rate limits

Every API response includes budget headers so your app (or agent) can self-regulate:

```text
x-recursiv-tier: free
x-recursiv-calls-remaining: 4847
x-recursiv-calls-limit: 5000
```

When you exceed your limit, the API returns `429 Too Many Requests` with a `Retry-After` header.

## Auto-retry

The SDK automatically retries requests that fail with 429 (rate limit) or 5xx (server error) using exponential backoff. It respects `Retry-After` headers.

```typescript
const r = new Recursiv({
  maxRetries: 3, // default is 2, set 0 to disable
});
```

If all retries are exhausted, the SDK throws a `RateLimitError` with:

* `retryAfter` — seconds to wait
* `upgradeUrl` — link to upgrade your plan (free tier only)

## Agent discovery

Agents can discover the full API surface and manage their own budget programmatically:

| Endpoint                         | Purpose                                               |
| -------------------------------- | ----------------------------------------------------- |
| `GET /api/v1/openapi.json`       | Complete OpenAPI specification                        |
| `GET /api/v1/llms.txt`           | Condensed reference optimized for LLM context windows |
| Budget headers on every response | Remaining quota so agents can plan ahead              |

## Anonymous sandbox limits

The anonymous sandbox (`POST /api/v1/sandbox/try`) is free with no signup:

* 10 executions per IP per day
* TypeScript, JavaScript, and Python supported
* 30-second execution timeout
* No persistent storage

## Enterprise

For teams that need dedicated infrastructure, compliance, or custom terms:

* Everything in Pro
* SOC 2 compliance + dedicated support
* SSO / SAML integration
* Custom SLAs and volume pricing
* Self-hosting option with full data control

Contact [hello@recursiv.io](mailto:hello@recursiv.io) for enterprise pricing.