Skip to content

HTTP 429 Too Many Requests With a Proxy: Pacing and Sessions

Slow down first, then find out what the limit is tied to, and only then think about the exit.

The 3 most common causes

The request rate or parallelism is above what the site allows. A loop without pauses, many threads at once or a retry that fires instantly all count. MDN describes limits as server-wide or per resource.

The limit is tied to something other than the IP: an account, an API key, a token or a cookie. Then every IP you use is counted against the same login, and a new exit changes nothing.

Too many requests leave through one IP. A fixed IP or a long session with many threads piles every request onto one address, so the limit is reached quickly even when each thread is slow.

What the limit is tied to and what to change

What the limit is tied to and what to change
ItemHow it shows upWhat helpsWhat does not
Per IP address429 starts after a burst and clears after a wait; another network is not limited.Lower the rate and the parallelism; honour Retry-After. Spread traffic across more connections only where the site allows that volume.Hammering the same URL harder; a proxy does not raise the allowance.
Per account, key or cookie429 follows the login or key on any IP.Cut the call count, cache answers, use the quota the key was given, read the API docs.Changing IP. For a logged-in account keep one fixed IP instead of switching.
Per page or endpointOnly one URL returns 429.Request that URL less often, or fetch it once and reuse the result.Retrying it in a tight loop.

Based on MDN (429 Too Many Requests, Retry-After) and the Cloudflare documentation for Error 1015, read on 16 November 2026. The table is a diagnostic guide, not a rule of every site.

Fix it in 6 steps

  1. Step 1: Read the answer

    Print the status line and the Retry-After header (curl line below). A number means seconds to wait; a date means wait until then.

    Expected result HTTP/2 429 with retry-after: 60 means wait 60 seconds.

  2. Step 2: Wait, do not retry at once

    Cloudflare warns that repeated retries in a short window may extend the block. Wait the Retry-After time, or at least a few minutes if none is sent.

  3. Step 3: Lower parallelism and add a pause

    Halve the number of threads or tasks, then put a fixed pause between requests. Raise the rate again only in small steps while 429 stays away.

  4. Step 4: Add backoff in the client

    Retry 429 with a growing delay and let the library honour Retry-After (code below). Cap the number of retries so a stuck job fails clearly.

  5. Step 5: Choose the exit mode on purpose

    On a rotating line every new connection gets a new IP, while a reused connection (keep-alive) keeps its IP. For a job that must look like one visitor, or a logged-in account, open a session or use a static IP instead.

  6. Step 6: Reduce what you request

    Cache pages that do not change, skip duplicates, and use the official API with its quota when the site has one. Fewer requests beat any exit setting.

Read the status and Retry-After

curl -sS -D - -o /dev/null --connect-timeout 10 -x http://LOGIN:PASSWORD@HOST:PORT https://example.com/ | grep -i -E '^(HTTP|retry-after)'

Example only: prints the status line and the Retry-After header if the site sends one.

Retry 429 with backoff through a proxy

import requests
from urllib3.util import Retry
from requests.adapters import HTTPAdapter

PROXY = "http://LOGIN:PASSWORD@HOST:PORT"

retry = Retry(
    total=5,
    backoff_factor=1,
    status_forcelist=[429, 503],
    respect_retry_after_header=True,
)

s = requests.Session()
s.proxies.update({"http": PROXY, "https": PROXY})
s.mount("https://", HTTPAdapter(max_retries=retry))
s.mount("http://", HTTPAdapter(max_retries=retry))

r = s.get("https://example.com/", timeout=(5, 30))
print(r.status_code, r.headers.get("Retry-After"))

Example only: replace LOGIN, PASSWORD, HOST and PORT. The retry behaviour was run against a local test server that answered 429 with Retry-After: 1. After the retries run out, requests raises RetryError. If HTTP_PROXY variables are set, pass proxies on each call.

If it still fails

429 comes back after every wait

Check Whether the limit is tied to your account or key rather than the IP.

Fix Reduce the total number of calls and read the API quota rules.

Stop when If it is a login or key limit, stop changing IPs; it will not help.

Cloudflare Error 1015 page

Check The limit is set by the site owner, not by Cloudflare or by the proxy.

Fix Wait, then resume at a lower rate.

Stop when If it persists after a long wait, contact the site owner or use their API.

429 only on some requests through the proxy

Check The connection mode: a new IP per connection or a session, and how many threads share it.

Fix Cap the threads per session and add the pause.

How to do it with Proxy302

In the order page choose Rotating residential, then HTTP (port 2222) or SOCKS5 (port 3333) on proxy.proxy302.com. Without a session every new connection gets a new IP and a reused connection keeps its IP. With a session the IP is fixed: the Budget line takes 1 to 120 minutes (10, 30, 60 and 120 are preset), and the Premium line usually keeps one IP while it is in use.

There is no limit on concurrent connections. That is a technical fact, not permission to exceed a site's limit, so keep your own pace.

Prices and limits

The Budget line costs US$1.50 per GB and the Premium line US$5 per GB; purchased traffic has no expiry, and 195+ countries can be targeted. There is no free trial; the smallest purchase is 1 GB.

FAQ

How long should I wait after a 429?

As long as Retry-After says, in seconds or until the given date. If the header is missing, start with a few minutes and lengthen the pause if 429 returns.

Does a rotating proxy fix a 429?

Not by itself, because a proxy does not raise the limit a site sets. If the limit is per IP and the site allows your volume, spreading connections can help; if it is tied to an account or key, the IP does not matter. Pace the requests first.

What is the difference between 429 and 403?

429 says too many requests in a given time and often clears after a wait. 403 says the server refuses the request and repeating it unchanged will fail the same way.

Should I use a sticky session or a rotating IP?

For one logged-in account or a multi-step flow use a session or a static IP. For many independent public pages a rotating line fits, at a pace the site allows.

Can too many threads on one proxy cause 429?

Yes. If a session or static IP carries every request, they all count toward one address. Cap the threads per IP and add a pause.

Reference links

6
  1. MDN: 429 Too Many Requests

    Official documentation · Confirmed on:

    Read on 16 November 2026: 429 means the client sent too many requests in a given amount of time (rate limiting); limits can be server-wide or per resource and are typically based on the IP address, but can be tied to users or authorized applications; a Retry-After header may say how long to wait.

  2. MDN: Retry-After header

    Official documentation · Confirmed on:

    Read on 16 November 2026: Retry-After carries either a number of seconds or an HTTP date; it is used with 429, 503 and redirect responses.

  3. Cloudflare docs: Error 1015 You are being rate limited

    Official documentation · Confirmed on:

    Read on 16 November 2026: Error 1015 means the site rate-limited you for sending too many requests in a short period; the advice is to wait and try later, because repeated retries within a short window may extend the block; the site owner sets the limit.

  4. urllib3 documentation: urllib3.util.Retry

    Official documentation · Confirmed on:

    Read on 16 November 2026: status_forcelist lists the statuses that force a retry; backoff_factor sleeps backoff_factor * 2 ** (retries so far) seconds, capped by backoff_max (default 120); respect_retry_after_header defaults to True and covers 413, 429 and 503. The example was run against a local test server that answered 429 with Retry-After: 1 twice: it waited and then returned 200, and raised RetryError when 429 never stopped.

  5. Python Requests documentation: Advanced Usage

    Official documentation · Confirmed on:

    Read on 16 November 2026: proxies are set with the proxies argument or session.proxies; proxy environment variables can override session.proxies, so pass proxies on each request when they are set; Retry is attached with HTTPAdapter(max_retries=...); set a timeout on requests.

  6. Proxy302 product facts and price list

    Product observation · Confirmed on:

    Product facts file of the site, read on 16 November 2026: rotating residential prices per GB, protocols and ports, countries; and the dynamic-line behaviour confirmed on 11 November 2026: a new IP for every new connection unless a session is opened, sessions of 1 to 120 minutes on the Budget line, no limit on concurrent connections.

Ready to try it?

Create an account, top up once, then pick rotating residential, static ISP or datacenter IPs and have one running in your browser within minutes.