What does a 429 error mean?

A 429 status means you have hit a rate limit — too many requests per minute, too many tokens per minute, or you have run out of quota. The fix is to slow down with exponential backoff, batch fewer requests, or upgrade your tier. Most SDKs include a Retry-After header telling you how long to wait.

Why am I getting a 401 unauthorized error?

A 401 means the API rejected your credentials. Usually the API key is missing, malformed, revoked, or belongs to the wrong organization or project. Regenerate the key, confirm it is loaded from your environment correctly, and check that the right organization or project header is set.

What causes a context length exceeded error?

This 400-class error fires when your prompt plus the requested completion is larger than the model's context window. Reduce the input (truncate or summarize history), lower max_tokens, or switch to a model with a bigger context window. The error message usually states the limit and how many tokens you sent.

Does this tool send my error to a server?

No. All matching and explanation happens locally in your browser. Your error text, which can contain request IDs or snippets of prompts, is never uploaded.

It says unknown error — what now?

If the decoder cannot match the error, copy the status code and the type or message field and search the provider's status page and error reference. A 500 or 503 is almost always a transient server-side issue you should retry with backoff rather than a bug in your code.

LLM API Error Decoder

LLM API errors are terse and provider-specific. This decoder takes the raw message or JSON body and explains, in plain English, what went wrong, the most likely fix, and what to check next — across OpenAI, Anthropic, Gemini and Mistral.

How it works

Paste the error. The tool looks for an HTTP status code and an error type / code field, matches them against a knowledge base of common LLM failure modes, and returns a cause, a recommended action and a category (auth, rate limit, request, server). Matching runs entirely in your browser — nothing is uploaded, so request IDs and prompt fragments stay private.

Common cases

401 / invalid_api_key — credentials missing, wrong, or revoked. Rotate the key and check the org/project header.
429 / rate_limit_exceeded — too many requests or tokens. Back off exponentially, honour Retry-After, or raise your tier.
400 / context_length_exceeded — prompt plus completion exceeds the window. Shorten input, lower max_tokens, or move to a larger model.
500 / 503 — transient provider-side issue. Retry with backoff; check the status page.

If the decoder cannot classify your error, lean on the status code: 4xx means your request needs fixing, 5xx means retry the provider.