generated: '2026-07-19' method: searched source: https://docs.ltx.io/errors docs: https://docs.ltx.io/errors format: custom envelope: shape: '{"type": "error", "error": {"type": "", "message": ""}}' discriminator_field: error.type schema: openapi/lightricks-ltx-openapi.yml#/components/schemas/Error notes: >- The same {type, message} shape appears in three places: v1 sync response bodies on failure, v2 async submit response bodies on failure, and the `error` field of a v2 job status payload whose `status` is `failed`. Error handling can key on `error.type` regardless of origin. error_types: - type: invalid_request_error status: 400 retryable: false meaning: Invalid request parameters. - type: authentication_error status: 401 retryable: false meaning: API key missing or invalid. remediation: 'Send `Authorization: Bearer YOUR_API_KEY`; verify the key in the developer console.' - type: insufficient_funds_error status: 402 retryable: false meaning: Not enough credits. remediation: Top up credits or enable auto top-up. docs: https://docs.ltx.io/auto-top-up - type: not_found_error status: 404 retryable: false meaning: Job doesn't exist or has expired. remediation: Job status and output URLs are retained for 24 hours after a terminal state. - type: content_filtered_error status: 422 retryable: false meaning: Content rejected by safety filters. - type: rate_limit_error status: 429 retryable: true meaning: Too many requests in a time window. remediation: Honor the `Retry-After` header; back off exponentially with jitter. - type: concurrency_limit_error status: 429 retryable: true meaning: Too many concurrent generations (default limit 2). remediation: Queue submissions or contact sales for a higher concurrency limit. - type: api_error status: 500 retryable: true meaning: Unexpected server error. - type: service_unavailable status: 503 retryable: true meaning: Service temporarily unavailable. - type: overloaded_error status: 529 retryable: true meaning: API temporarily overloaded. remediation: Retry after a short delay. retry_guidance: >- For retryable errors use exponential backoff with jitter — a random delay of up to 50% of the backoff interval prevents thundering-herd retries. On 429, honor the Retry-After header if present. support: request_id_header: x-request-id docs: https://docs.ltx.io/debugging