What happens when an RPC endpoint hits its daily request cap?
When an nodes/rpc-endpoint-basics/">RPC endpoint hits its daily request cap, the provider stops honoring new requests from your account until the cap resets - typically at midnight UTC. You will receive HTTP 429 (Too Many Requests) or 403 (Forbidden) error responses for all subsequent calls. Transactions, queries, and smart contract interactions fail silently or return these errors, depending on your client configuration.
How providers enforce the cap
Public and free-tier RPC providers like Infura, Alchemy, and QuickNode track usage per API key or IP address. Each request - whether a simple eth_blockNumber or a complex eth_call - counts toward a daily limit. The cap is a hard ceiling: once reached, the provider rejects requests until the next billing cycle starts.
The exact behavior depends on the provider:
- Infura: Returns HTTP 403 with an error message like "Project ID exceeds its daily request limit." The header
X-RateLimit-Remainingdrops to zero before the cap is hit. - Alchemy: Returns HTTP 429 with a JSON-RPC error code -32000. Their dashboard shows remaining compute units (CU) rather than raw requests.
- QuickNode: Returns HTTP 429 and a
too_many_requestserror. They also provide real-time usage in their dashboard.
Some providers offer a grace period - a handful of extra requests past the cap - but this is not guaranteed and varies by plan. Most simply cut off service immediately.
What fails and what doesn't
When the cap is hit, all RPC methods fail. This includes:
- Reading balances, token metadata, or transaction history
- Sending transactions (they never reach the mempool)
- Calling
eth_estimateGasoreth_call - Subscribing to events via WebSocket
- Using
eth_getLogsfor historical queries
Cached data or local copies of state are unaffected. If your application caches block headers or token prices, those continue to serve until expired. But any live interaction with the blockchain stops entirely.
How to Detect You've Hit the Cap
Your application should monitor for these signs:
- HTTP status codes: 429 or 403 from the RPC endpoint
- Error messages: Look for strings like "rate limit," "daily limit," "too many requests," or "exceeded compute units"
- Consistent failures: All requests fail, not just heavy ones like
eth_getLogswith large ranges - Dashboard warnings: Most providers send email alerts or show a "cap reached" indicator in their admin panel
If only some requests fail (e.g., eth_call works but eth_getLogs returns an error), you likely hit a per-method rate limit, not the daily cap. See the site's coverage of "How to fix eth_getLogs query returned more than 10000 results" for that issue.
What to Do When the Cap Is Reached
Immediate fixes
- Switch to a backup RPC endpoint - If you have a multi-provider failover setup (covered in "How to set up multi-provider RPC failover for Ethereum"), the fallback kicks in automatically. Otherwise, manually update your RPC URL to a different provider or a secondary API key.
- Use a private RPC endpoint - If you run your own node, point your application there. Self-hosted nodes have no daily cap (only hardware limits).
- Wait for the reset - Most free tiers reset at midnight UTC. If your use case is non-urgent, simply pause requests until the next day.
Long-term solutions
- Upgrade your plan - Free tiers typically allow 100,000 - 500,000 requests per day. Paid plans raise or remove the cap. Check your provider's pricing page for current limits.
- Reduce request volume - Batch queries, cache responses, and use
eth_getLogswith narrower block ranges. See "How to fix eth_getLogs query returned more than 10000 results" for efficient querying. - Run your own node - This eliminates provider caps entirely. See "RPC endpoints and nodes" and "Full node vs archive node" for guidance on whether this fits your needs.
A note on compute units
Some providers (notably Alchemy) use compute units (CU) instead of raw request count. A simple eth_blockNumber might cost 1 CU, while a complex eth_call on a heavy contract could cost 20 - 50 CU. If you hit the daily CU cap, the same errors occur. Monitor your CU usage in the provider dashboard, not just request count.
Why this happens more often than you expect
Daily caps are designed to prevent abuse and share free-tier resources fairly. But legitimate applications often hit them due to:
- Polling loops in dApps that check balances every few seconds
- Automated scripts or bots running without rate limiting
- Multiple applications sharing one API key
- Sudden traffic spikes (e.g., a popular NFT mint)
If you consistently hit the cap, it's a sign your application needs better request management or a higher-tier plan.
Not financial advice. starname.me publishes market data and general information about digital assets. Crypto assets are volatile and you can lose everything you put in. Nothing here is a recommendation to buy, sell or hold, and we make no price predictions.
Prices are sourced from third parties and may be delayed or wrong. Verify anything you intend to act on against a primary source.