What 529 Overloaded actually means
As of 2026-10-20, 529 Overloaded means the Claude API is temporarily at capacity: the official error reference states this is a capacity condition across all users, not your usage limit, and it doesn't count against your quota. This recheck also fills in the retry defaults the official docs document: CLAUDE_CODE_MAX_RETRIES defaults to 10 (capped at 15 since 2.1.186), the base delay for 529 retries defaults to 500 ms (CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS, raisable to 32000), and the backoff ceiling defaults to 32000 ms (CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS, raisable to 64000). Changing your code will not help; changing your retry strategy will.
Updated 2026-10-20
One key for Claude, GPT and leading Chinese models, billed per token, up and running in minutes.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four points
Vendor-side congestion
Returned when server-side capacity is tight. Unrelated to your key's quota, your parameters, or your request body. The official error reference states that a 529 is a capacity condition across all users, so it points at the vendor's side, not yours.
You are being rate limited
This is the one that means you exceeded your own rate or quota. 529 and 429 need different handling — do not share a code path. One detail in the official docs is easy to miss: when a 429 reports a spend limit or exhausted usage credits, Claude Code fails at once instead of retrying, the opposite of how a 529 is handled.
The only effective client-side move
Exponential backoff with jitter. Retrying immediately worsens the congestion and gets you rate limited sooner. The official docs give the default base delay for 529 retries as 500 ms (CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS) and the default backoff ceiling as 32000 ms (CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS).
Structural mitigation
When the same work can land on a different model family, one congested vendor stops meaning downtime. The cost is accepting model differences.
How 529 differs from 429 and 503
As of 2026-10-20, 529 means the server is overloaded right now — a capacity problem, usually temporary; the official error reference states this is a capacity condition across all users, not your usage limit, and it doesn't count against your quota. 429 means you hit your own rate or quota ceiling — a quota problem; the official docs also state that a 429 reporting a spend limit or exhausted usage credits makes Claude Code fail at once instead of retrying. 503 generally means the service is unavailable or under maintenance.
None of them means your request was malformed, but the handling differs: 529 calls for backoff and retry, 429 calls for slowing down or raising limits, 503 calls for waiting and watching the status page. Collapsing all three into one catch block is the most common mistake in error handling.
Why discussion picks up periodically
Whenever a new model ships or a large migration happens, vendor capacity tightens for a while and 529 discussion rises with it. These waves usually ease as capacity is added — but for someone on a deadline, waiting for a vendor to expand is not an available answer. The official status page recorded two checkable entries in 2026-10: on 2026-10-07 an issue with spend limits left some organizations incorrectly paused for having reached their spend limit, so requests were refused on the Claude API, Claude Code and the other Claude surfaces, and the vendor marked it resolved; on 2026-10-13 an incident on platform.claude.com about usage data explicitly said that API requests were not affected.
Both entries concern spend limits and usage data, not capacity. Claude Code also shipped two versions in quick succession on 2026-10-08 and 2026-10-09 (2.1.295 and 2.1.296), adding a cap on the total wait in unattended retry mode and a configurable 529 backoff ceiling; the latest version is 2.1.304, released 2026-10-19.
Diagnostic order
Confirm the status code really is 529 and not 429. The response bodies differ; count them separately in your logs.
Check the vendor status page. If it is a broad incident, no client-side change is more than mitigation. The official status page lists Claude API (api.anthropic.com) and Claude Code as two separate components, so check which one is affected; as of 2026-10-20 every component on the page reads operational.
Check whether your own retry logic is amplifying the congestion: no jitter, no ceiling, and immediate retry all amplify. On Claude Code, since version 2.1.292 (2026-10-06) you can lengthen the base delay of 529 retries with CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS; since 2.1.295 (2026-10-08) CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS caps the total time spent waiting out 429 and 529 in unattended retry mode; since 2.1.296 (2026-10-09) CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS lets you raise the backoff ceiling.
Before retrying, work out which code you actually got
What is confirmed
A 529 signals server-side capacity congestion and is unrelated to request content — this semantics is stated in vendor documentation. Claude Code's official error reference (checked 2026-10-20) shows the message “API Error: Repeated 529 Overloaded errors. The API is at capacity — this is usually temporary.” and states: “A 529 is not your usage limit and doesn't count against your quota.” It adds that capacity is tracked per model, so you can run /model and switch to a different model to keep working, and that a 529 is a capacity condition across all users.
The same error reference gives the retry defaults: CLAUDE_CODE_MAX_RETRIES defaults to 10 (capped at 15 since 2.1.186), CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS defaults to 500 ms (raisable to 32000), CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS defaults to 32000 ms (raisable to 64000), and CLAUDE_CODE_RETRY_WATCHDOG set to 1 retries 429 and 529 capacity errors indefinitely.
The Claude Code CHANGELOG records that from 2.1.286 a failing model call sends at most 14 requests with the default retry settings, that 2.1.292 added CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS, that 2.1.295 added CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS, and that 2.1.296 added CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS. Exponential backoff with jitter is a widely validated client-side response.
Do not treat as fact
For a 529 the error reference gives three actionable steps: check the status page for capacity notices, try again in a few minutes, and run /model to switch to a different model; it also states that a 529 is a capacity condition across all users and that capacity is tracked per model. Treating 529 as the vendor quietly throttling heavy users has no basis, and conflating it with 429 sends you to the wrong fix.
Two responses
Client-side backoff only
Cheap to implement and measurably reduces failures. But while the vendor stays congested, all you can do is wait.
Let work land on multiple model families
One congested provider stops meaning downtime. The cost is accepting different model habits, and every route has its own failure modes. Multiple routes are mitigation, not elimination: the switching mechanism itself can fail as well.
Writing retries correctly
Three points. First, exponential backoff rather than a fixed interval: wait a second, then double each time. Second, add random jitter — if every client backs off by the same amount they retry in a synchronised wave, which prolongs the congestion. Third, set ceilings on both retry count and total wait, or a single congestion event will let your task queue grow without bound.
Separately, when a streaming request breaks mid-way, do not blindly retry from the start; first check whether what you already received can be continued. If you use Claude Code: its documentation says it retries transient failures up to 10 times with exponential backoff before showing an error (rechecked 2026-10-20); version 2.1.286 (GitHub release 2026-09-30) changed how retries are counted — the changelog means that one limit now covers a whole model call, so with the default retry settings a failing call sends at most 14 requests; version 2.1.292 (GitHub release 2026-10-06) added the environment variable CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS, which sets a longer base delay for the backoff when retrying an overloaded 529 request; version 2.1.295 (GitHub release 2026-10-08) added CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS, which caps how long unattended retry mode waits out 429 and 529; version 2.1.296 (GitHub release 2026-10-09) added CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS to raise the ceiling of the 529 backoff.
The official error reference now states the defaults: CLAUDE_CODE_MAX_RETRIES defaults to 10 (capped at 15 since 2.1.186), CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS defaults to 500 ms (raisable to 32000), CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS defaults to 32000 ms (raisable to 64000), and CLAUDE_CODE_RETRY_WATCHDOG set to 1 retries 429 and 529 capacity errors indefinitely.
What QCode can and cannot do here
What it can do: one key lets you switch to another model family during congestion instead of waiting for a single model to recover. What it cannot do: we are not the model vendor and cannot change the vendor's server capacity, and requests through QCode can fail too. What we offer is the option to switch model families, not a promise of flawless service.
FAQ
Should I change my request parameters when I get a 529?
No. 529 is unrelated to request content; changing max_tokens, model parameters, or trimming the prompt will not make it go away. Change the retry strategy instead.
Can I use the same retry logic for 529 and 429?
Not advisable. 529 warrants backing off and retrying the same request; 429 means you need to lower your send rate or raise your limits, and blind retries keep tripping the limiter.
How long should I back off?
A common pattern is one second, doubling each attempt, with random jitter and ceilings on both retry count and total wait. The exact numbers depend on how much latency your task tolerates. In Claude Code (official error reference and CHANGELOG as of 2026-10-20): CLAUDE_CODE_MAX_RETRIES sets the number of retry attempts, default 10 (capped at 15 since 2.1.186); since 2.1.286 a failing model call sends at most 14 requests with the default retry settings; since 2.1.292 CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS lets you set a longer base delay for 529 retries, and the official docs give 500 ms as the default, raisable to 32000; since 2.1.295 CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS caps the total time spent waiting out 429 and 529 in unattended retry mode; since 2.1.296 CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS lets you raise the backoff ceiling, and the official docs give 32000 ms as the default, raisable to 64000; setting CLAUDE_CODE_RETRY_WATCHDOG to 1 retries 429 and 529 capacity errors indefinitely.
What if a streaming request hits 529 mid-way?
First check whether what you already received is usable. Continue if you can; retry the whole thing only if you cannot. Blindly restarting duplicates billing and adds to the congestion.
Does QCode get 529s?
It can. Congestion on the vendor side is real and we are not immune to it. What we can do is let the same work move to another model family — not promise that failures will not happen.
What does 529 overloaded mean, and why do I get it when my quota is unchanged?
529 Overloaded is a capacity state on the vendor side: the official error reference states that a 529 is a capacity condition across all users, not your usage limit, and it doesn't count against your quota, so an unchanged quota is expected. Capacity is tracked per model, so you can run /model and switch to a different model to keep working; the official advice is to try again in a few minutes and to check the vendor status page for capacity notices. Raising your quota or changing parameters does not address it. Multiple routes lower the probability that one outage stops your work, but each route has its own failure modes and the switching mechanism itself can fail: it is mitigation, not elimination.
Sources
Status-code semantics follow each vendor's official API documentation. Claude Code's retry behaviour and CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MS, CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS and CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS come from Anthropic's Claude Code CHANGELOG (the 2.1.286, 2.1.292, 2.1.295 and 2.1.296 entries), the GitHub release pages of those versions, the Claude Code error reference and environment-variable documentation, and the official status page with its component list and incident records, all fetched and checked on 2026-10-20. This page deliberately cites no third-party availability percentages — such figures move over time and we have no authoritative independent measurement to point to.
When one route is congested, take another
One key across multiple model families — switch and keep going when a route is under pressure.
Related reading
Claude Code FAQ and error handling
Answers to common questions while using Claude Code, covering onboarding, pricing quotas and troubleshooting.
Claude Code usage limits
Tell the 5-hour window apart from the weekly limit, and what to do once the usage limit is reached.
Multi-model routing
A decision matrix for choosing and switching between model families with one key.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Status-code semantics here follow each vendor's official documentation and may change between versions; the Claude Code details were checked on 2026-10-20 against the official CHANGELOG and documentation, which take precedence. Nothing on this page is an availability commitment.