Error Codes

Import the API collection into Bruno, Postman, or Insomnia to try all examples locally.

Error Response Format

All error responses share a consistent JSON structure:

{
  "code": "RATE_LIMIT",
  "what": "Rate limit exceeded",
  "why": "You've exceeded the free playground request limit",
  "action": "Wait for reset, or upgrade to SAP Generative AI Hub for unlimited requests"
}

Field	Description
`code`	Machine-readable error code (see tables below)
`what`	Short human-readable title
`why`	Explanation of what went wrong
`action`	Suggested next step

HTTP Status Codes

Code	Meaning	Description
400	Bad Request	Invalid data format or validation error
401	Unauthorized	Invalid or missing API token
429	Too Many Requests	Rate or burst limit exceeded (includes `Retry-After`) — see Limits & Quotas
500	Internal Server Error	Contact support
502	Bad Gateway	Remote URL fetch failed (see `DS_FETCH_FAILED`)
503	Service Unavailable	Server under high load. Retry after `Retry-After` header
504	Gateway Timeout	Remote fetch timeout (see `DS_FETCH_TIMEOUT`)

Application Error Codes

Code	HTTP	Description
`PRED_QUERY_ROWS_MIN`	400	No query rows — add at least one `[PREDICT]` marker
`PRED_CONTEXT_MIN`	400	Too few context rows (minimum 2)
`PRED_DATE_FORMAT`	400	Date column values not parseable — see Date Formats
`PRED_EXPLANATIONS_INVALID`	400	Invalid `explanations` value — field path and reason included in message
`DS_FETCH_FAILED`	502	Remote URL returned error (404, 403, etc.) or response not parseable as CSV
`DS_FETCH_TIMEOUT`	504	Remote fetch exceeded 30-second timeout
`DS_URL_BLOCKED`	400	URL blocked (private IP, redirect, or non-HTTPS)

Free Playground Limit Errors

These errors apply only to the free playground. They do not exist in SAP Generative AI Hub.

Code	HTTP	Description
`PRED_CONTEXT_MAX`	400	Too many context rows for the free playground (maximum 100,000)
`PRED_QUERY_ROWS_MAX`	400	Too many `[PREDICT]` rows for the free playground (maximum 512)
`PRED_COLS_MAX`	400	Too many target columns for the free playground (maximum 10)
`DS_FILE_TOO_LARGE`	400	Remote file exceeds the free playground size limit (maximum 100 MB)
`DS_TOO_MANY_ROWS`	400	Remote dataset exceeds the free playground row limit (maximum 100,000)
`RATE_LIMIT`	429	Free playground hourly request budget exceeded
`RATE_LIMIT_BURST`	429	Requests sent too quickly — burst limit hit (retry after a few seconds)

Generative AI hub

No rate limits, no dataset caps, enterprise SLAs.

Unlimited API calls – faster performance, more context
Enterprise security – SOC 2, ISO 42001
Commercial license – production SLAs, 24/7 support

Learn More

Rate Limit Headers

When rate-limited (429), the response includes additional headers:

X-RateLimit-Limit — total allowed requests
X-RateLimit-Remaining — remaining requests in window
X-RateLimit-Reset — Unix timestamp when window resets
Retry-After — seconds until next request is allowed

A 429 is returned as either RATE_LIMIT_BURST (requests sent too quickly) or RATE_LIMIT (hourly budget exhausted) — see the table above. Wait for the Retry-After duration before retrying. For exact burst and hourly values, see Limits & Quotas.