reference

Errors and limits

Errors use the OpenAI envelope, so existing client error handling works:

{
  "error": {
    "message": "Your credit balance is empty. Add credits at ...",
    "type": "insufficient_quota",
    "param": null,
    "code": "insufficient_credit"
  }
}

Codes

status	code	meaning	retry?
400	`invalid_json`	The body was not valid JSON.	No
400	`missing_text`	`/v1/restyle` was called without a non-empty `text`.	No
400	`missing_question`	`/v1/fanout` was called without a non-empty `question`.	No
400	`input_too_large`	Over the input bound for the model you named: 6,000 characters for a restyle, which you send a section at a time, or 2000 for a fan-out question. The message names the length you sent.	No
400	`missing_state`	`/v1/systemone` was called without a `state`.	No
400	`invalid_questions`	A question in a `/v1/systemone` request does not follow the schema: an unknown type, missing instructions, or criteria of the wrong shape or size. `param` names the question.	No
400	`input_too_long`	Over the 4,096-token context: a restyle block with no paragraph break in it, or a `/v1/systemone` state and questions that are too long together. Split it and send it again.	No
400	`wrong_endpoint_for_model`	The model named does not do the work this endpoint does — a rewrite model at `/v1/fanout`. Served badly it would return a plausible-looking answer, so it is refused.	No
401	`invalid_api_key`	Missing, malformed, unknown or revoked key.	No
402	`insufficient_credit`	The account balance is empty. Add credit, or turn on auto top-up.	After topping up
402	`key_spend_limit_reached`	This key hit its lifetime cap. Other keys are unaffected.	After raising it
403	`model_not_permitted`	The key’s allowlist does not include this model.	No
404	`model_not_found`	No such model id.	No
404	`model_not_available`	The model is in the catalogue but not serving yet.	No
502	`empty_generation`	A fan-out produced nothing usable for that question, so there is no list to return.	Once, then rephrase
429	`capacity_saturated`	Our capacity is full, or ran out while your request was running. When extra capacity is starting, the message says so and `Retry-After` is 30 seconds; otherwise it is 5. Nothing is charged, and a restyle never comes back half done: it is either complete or this error.	Yes, after `Retry-After`
502	`upstream_error`	The model returned an error.	Yes, with backoff
502	`upstream_unreachable`	The model could not be reached.	Yes, with backoff
503	`no_capacity`	No capacity is available. A problem on our side, not with your request.	Yes, with backoff
504	`timeout`	The model took too long.	Yes

In short: retry 429 after the Retry-After header, and 5xx with exponential backoff and jitter — with the exception of empty_generation, where the model ran and simply produced nothing usable, so one retry is worth it and a loop is not. Do not retry 4xx, nothing about the request will have changed.

Asking about a specific request

Every API response carries a request ID: as id in the body and as the X-Request-Id header, on errors as well as successes. Chat completions use their chatcmpl- id. Keep it with anything you might want to ask us about, and quote it when you do.

You can look a request up yourself, with any key on the same account:

curl https://meragpt.com/v1/requests/req_4f1c9a0b2e7d6c5a3b8e9f01   -H "Authorization: Bearer $MERAGPT_API_KEY"

{
  "id": "req_4f1c9a0b2e7d6c5a3b8e9f01",
  "object": "request",
  "model": "restyler-1",
  "status": "succeeded",
  "latency_ms": 4210,
  "usage": { "input_tokens": 412, "output_tokens": 377, "cost_usd": "$0.000077" },
  "detail": {
    "rewritten": 3,
    "total": 4,
    "blocks": [
      { "index": 2, "words": 61, "status": "unchanged", "reason": "invented" }
    ]
  }
}

It tells you what happened — which paragraphs were rewritten or left unchanged and why, tokens, cost, latency and any error — but never what was said: we do not store the text you send or the text we return. See our privacy policy.

Failed requests are not charged

A request that fails before the model generates anything costs nothing. It is still recorded, at zero cost, so a failing integration is visible on your usage dashboard instead of only in our logs.

Limits

limit	value
Request body, restyle	6,000 characters per request
Single block, restyle	~1,150 words
Question, fan-out	2000 characters
Brands and category, fan-out	Up to 12 brands; each brand and the category up to 100 characters
Queries returned, fan-out	Up to 6, after near-duplicates are dropped
Context window	4,096 tokens, prompt and completion together
Request timeout	180 seconds per block
Playground, signed out	10 runs a day per IP, 2,000 characters each

Per-key concurrency limits apply so that one caller cannot occupy the whole fleet. If you are planning something high-volume, email hello@meragpt.com first and we will make sure the capacity is there.