# Limits and pricing

> Keyless capabilities, rate-limit headers, parameter bounds and credit packages.

Source page: https://docs.replynodes.com/docs/production/limits

# Limits and pricing

This page collects the limits that are published for ReplyNodes. A value that is not listed here is not published yet, so do not assume one. The [API reference](https://docs.replynodes.com/reference) is authoritative for parameter bounds.

## Keyless capabilities

Some capabilities work without a key, up to a per-caller daily allowance. The numbers are not published and can change, so read the `X-RateLimit-*` headers on each response instead of hard-coding a value.

Keyless: url-to-markdown, brand-kit, brand-logo, app-store-api, google-play-api, reddit-api, youtube-api, web-search, web-scraping, company-research, competitor-research and `GET /v1/jobs/company` (hiring signals). Each capability has its own bucket. When a bucket is used up, the API returns `429 anonymous_limit_reached` with an `error.next_action` that explains how to get a key. See [Errors](https://docs.replynodes.com/docs/production/errors).

Keys and OAuth move a call to the authenticated path. A restricted agent key (before email verification) uses its own daily allowance for the keyless capabilities and never spends credits. See [Authentication](https://docs.replynodes.com/docs/auth).

## Rate-limit headers

Keyless responses carry these headers:

| Header | Meaning |
| --- | --- |
| `X-RateLimit-Limit` | The limit for the bucket |
| `X-RateLimit-Remaining` | Units left in the current window |
| `X-RateLimit-Reset` | When the window resets, in Unix seconds (the next UTC midnight for daily buckets) |
| `Retry-After` | Seconds to wait. Sent with `429` |
| `X-Request-Id` | The request id, also returned as `meta.request_id` |

On the free Markdown endpoint, every call spends one unit, including `HEAD` requests.

## Per-endpoint limits

| Area | Limit |
| --- | --- |
| Scrape `max_chars` | 1 to 1,048,576 Unicode code points |
| Crawl `max_pages` / `max_depth` | 1 to 50 pages and 1 to 3 levels; larger values are clamped |
| Markdown selectors | At most 10 entries and 512 bytes per header |
| Free Markdown request | Keep the connection open for up to 33 seconds |
| Search `q` | 1 to 512 bytes |
| Search `limit` / `page` | `limit` 1 to 50; `page` 1 or 2 |
| Competitor research | 1 to 5 competitor domains per request |

Pagination and result-count bounds for each source are in its guide: [Reddit](https://docs.replynodes.com/docs/guides/reddit), [YouTube](https://docs.replynodes.com/docs/guides/youtube), [App Store](https://docs.replynodes.com/docs/guides/app-store), [Google Play](https://docs.replynodes.com/docs/guides/google-play) and [Search](https://docs.replynodes.com/docs/guides/search).

## Packages and credits

Authenticated calls, including MCP calls, use prepaid package credits.

| Package | Included credits | Price |
| --- | ---: | ---: |
| Free | 500 one-time credits | Free |
| Developer | 10,000 credits per month | $19 per month |
| Pro | 100,000 credits per month | $99 per month |

The credit cost of each call comes from the capability and pricing contract, and a client cannot set or override it. Plans are on the [pricing page](https://replynodes.com/pricing), and billing is managed on the [billing page](https://platform.replynodes.com/billing). When a workspace has no credits left, credit-metered calls return `402 topup_required`. Only a human can add credits.

## Not published yet

These values are not published, so do not rely on a number:

- Rate-limit and quota numbers, for keyless and authenticated calls. Use the response headers.
- Service-level agreements, data freshness and data retention periods.
- Per-operation credit prices in these docs. Use the [pricing page](https://replynodes.com/pricing).

Public data can be stale, removed or unavailable upstream. Surface that uncertainty to your users instead of promising freshness or completeness.
