langgenius/dify · error · ProviderQuotaExceededError

provider_quota_exceeded

provider_quota_exceeded

Error message

Your quota for Dify Hosted Model Provider has been exhausted. Please go to Settings -> Model Provider to complete your own provider credentials.

What it means

ProviderQuotaExceededError (HTTP 400, code provider_quota_exceeded) raised in DatasetDocumentListApi.post when the underlying save/indexing call throws QuotaExceededError — the Dify-hosted (trial) model provider's usage quota has been exhausted. The default message instructs the user to supply their own provider credentials.

Source

Thrown at api/controllers/console/datasets/datasets_document.py:571

        knowledge_config = KnowledgeConfig.model_validate(console_ns.payload or {})

        if not dataset.indexing_technique and not knowledge_config.indexing_technique:
            raise ValueError("indexing_technique is required.")

        # validate args
        DocumentService.document_create_args_validate(knowledge_config)

        try:
            documents, batch = DocumentService.save_document_with_dataset_id(
                dataset, knowledge_config, current_user, session=session
            )
            dataset = DatasetService.get_dataset(dataset_id_str, session)

        except ProviderTokenNotInitError as ex:
            raise ProviderNotInitializeError(ex.description)
        except QuotaExceededError:
            raise ProviderQuotaExceededError()
        except ModelCurrentlyNotSupportError:
            raise ProviderModelCurrentlyNotSupportError()

        return dump_response(
            DatasetAndDocumentResponse,
            {"dataset": dataset, "documents": document_responses(documents, session=session), "batch": batch},
        )

    @setup_required
    @login_required
    @account_initialization_required
    @console_ns.response(204, "Documents deleted successfully")
    @with_current_user
    @with_current_tenant_id
    @rbac_permission_required(RBACResourceScope.DATASET, RBACPermission.DATASET_EDIT)
    @with_session
    def delete(
        self,

View on GitHub (pinned to ef8544b173)

Solutions

  1. Add your own provider credentials in Settings -> Model Provider (OpenAI/Anthropic/etc.) and switch the dataset's embedding model to use them.
  2. Reduce the indexing load (smaller documents, fewer segments) to stay within the hosted quota if you must use it.
  3. Contact Dify sales/support to increase the hosted quota if applicable.

Example fix

// before — hosted trial quota exhausted
POST /console/api/datasets/<id>/documents  →  400 provider_quota_exceeded

// after — use own provider
// Settings → Model Provider → add own OpenAI key, select as embedding model provider
POST /console/api/datasets/<id>/documents  →  200
Defensive patterns

Strategy: try-catch

Validate before calling

# No code-side quota check is authoritative — the provider enforces it.
# Best pre-check: track local usage against a budget below the hosted cap.
def likely_within_quota(estimated_tokens: int, used: int, cap: int) -> bool:
    return used + estimated_tokens < cap

if not likely_within_quota(est_tokens, tenant_used, hosted_cap):
    return error("Switch to your own provider credentials to avoid quota exhaustion.")

Type guard

def has_own_provider_credentials(tenant_id: str) -> bool:
    """True when the tenant is NOT solely dependent on the hosted trial."""
    # pseudocode: inspect provider rows for a non-hosted, valid-credential provider
    return any(not p.is_hosted_trial and p.has_credentials for p in list_providers(tenant_id))

Try / catch

from core.errors.error import QuotaExceededError
from controllers.console.app.error import ProviderQuotaExceededError

try:
    documents, batch = DocumentService.save_document_with_dataset_id(...)
except QuotaExceededError:
    raise ProviderQuotaExceededError()

Prevention

When it happens

Trigger: POST /console/api/datasets/<dataset_id>/documents (typically high_quality indexing) using the Dify-hosted trial OpenAI provider after the tenant's free token quota ran out. The embedding/LLM call during indexing is rejected upstream as over-quota and re-raised here.

Common situations: Cloud/SaaS trial tenant relying on Dify's hosted OpenAI hit the free-tier cap; a large document batch consumed the remaining quota mid-index; long-time tenant that never switched to its own API key.

Related errors

AI-assisted analysis of langgenius/dify@ef8544b173 (2026-08-12). Data as JSON: /api/errors/02a074d49e101ae2. Report an issue: GitHub.