Troubleshoot file processing

Find the message you actually saw, read what it means, and apply the fix. Guessing at causes from a vague symptom wastes more time than anything else on this page.

Problems with the file you uploaded

"This file type is not supported"

You uploaded something the tool cannot decode. The usual cause is a file whose name lies about its contents — a WebP or HEIC image renamed to .jpg by a downloader or a chat app.

Open the original in the application that created it and export a supported format. Renaming will never help: the extension is a label, and the decoder reads the bytes. If you are not sure what you really have, run Inspect images and check the reported format. Password-protected documents are a separate case — see the PDF section below, since an encrypted file is rejected rather than merely unsupported.

"The result is larger than your target"

The job finished, but the output missed the byte ceiling. Work through these in order:

1. Check whether the target is even achievable. Some files cannot be compressed below a threshold without destroying content. Compare your result with the target: a 5 MB scan will never reach 200,000 bytes without becoming unreadable.

2. Check whether you are aiming at the wrong number. This is the trap worth knowing by heart:

   The form says	It may mean	Actual bytes
   200 KB	200 × 1000	200,000
   200 KB	200 × 1024 (KiB)	204,800
   2 MB	2 × 1000²	2,000,000
   2 MB	2 × 1024² (MiB)	2,097,152

   A form that accepts "200 KB" but enforces 204,800 bytes is a 4,800-byte gap, which is small in absolute terms and decisive in practice. Our default target is the stricter 200,000 bytes so that both readings pass.

3. Reduce dimensions rather than quality alone. For a photo, halving the pixel width cuts the data to roughly a quarter and usually looks better than aggressive quality reduction at full size.

4. For a scanned PDF, use deep compression. Page images have to be recompressed, which means visible detail loss. Zoom in on the smallest text afterward and decide whether the trade is acceptable.

If an over-target file is still useful, best-effort mode can return one — but read its warning rather than treating the download button as proof of compliance. Strict mode refuses the output unless the rule passes.

"This PDF was rejected"

Encrypted, password-protected, attachment-bearing, or active-content PDFs are refused. There is no override. Export a plain PDF from the application that created it, if you have the right to do so.

Two related cases that surprise people:

• A signed PDF is not rejected, but rewriting it invalidates the signature. Keep the signed original; work on an unsigned copy.
• Wrong page count. If a merge or split fails on page count, remember that limits apply to the output: merging two 60-page PDFs makes 120 pages, which is over the limit even though both inputs were fine. The numbers are in formats and limits.

AttachReady cannot edit PDF text, repair a corrupt file, or sign documents. Those features do not exist, so no error message will lead you to them.

Problems with your allowance or access

"You have reached your daily limit"

Wait for the reset or sign in for the larger account allowance.

• The reset happens at 00:00 UTC, not local midnight. Convert to your own time zone before assuming the wait is longer than it is.
• Failed attempts count too. Retrying the same malformed file five times spends five attempts. Fix the cause before retrying.
• Guest allowances are also shared per network — 30 per source address — so a busy office or campus connection can exhaust the pool before you run a single job.

Current numbers: formats and limits.

"This file has expired"

The 24-hour retention window closed. Expired files cannot be recovered, and there is no extension available. Re-upload your original and run a new task. Download results promptly — the full retention behaviour is in file privacy.

"My download link does not work"

Check expiry first, then ownership. Account artifacts are scoped to the account that created them, so a link from one account will not open in another, and a link generated for an API key belongs to that key's account. Link sharing across accounts is not supported.

Problems with the API

API errors: 401, 403 and 429

Status	Meaning	Fix
401	The key is missing, malformed or revoked	Send Authorization: Bearer <key>. Check the key still exists at /settings/apikeys; revoked keys fail immediately.
403	The key is valid but the operation is not allowed	Check the tool's required auth level, your plan entitlement, and whether the request targets another account's file.
429	Rate limit hit	Each key defaults to 120 requests per minute. Back off and retry; the daily processing allowance is separate and shared across all keys on the account.

Never put an API key in a browser bundle, a public repository or a client-side environment variable. The API reference covers key creation, upload and the job lifecycle.

"My job is still running"

Async jobs need polling, and polling too aggressively causes its own problems. Query the returned job endpoint with bounded backoff rather than in a tight loop, and stop when the status is terminal.

If the connection dropped mid-request, retry with the same idempotency key. A new key creates a duplicate job that consumes another allowance and may produce a second artifact.

Ask for help without sending private files

Email support@attachready.com with the tool URL, roughly when it happened, the exact error message or code, your browser version, and the file's format and approximate size. A job ID helps if you have one.

Do not send API keys, passwords, identity documents or confidential attachments. A minimal, non-sensitive sample that reproduces the problem is both more useful and less risky than your original material. The upload error guide has a longer pre-submission checklist, and the MCP guide covers tool discovery for AI clients.