Model Builder Errors

The Model Builder service trains predictive AI models on your Data Cloud data. Errors from this service typically indicate that a training job request was invalid or missing required parameters, that a job or model artifact couldn’t be found, or that a platform dependency the service relies on was temporarily unavailable.

A model artifact referenced by the request couldn’t be found. This is a non-retriable error — the artifact won’t appear until a successful training job produces it.

This error occurs when:

• The artifact identifier is mistyped or refers to an artifact that was never produced.
• The job that should have produced the artifact failed before completing, or is still running.
• The artifact was deleted or removed by a retention policy.
• The artifact belongs to a different org or environment than the one being queried.

To resolve this error:

1. Confirm the artifact identifier matches one returned by a completed training job.
2. Verify the originating job finished successfully — artifacts are only available after the producing job reaches a successful terminal state.
3. Confirm you’re using credentials for the org that owns the artifact.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_ARTIFACT_NOT_FOUND
• The artifact identifier and the parent job ID, if known

The Model Builder service couldn’t reach its database to complete the request. This is a temporary platform-side issue — wait a moment and try again.

This error occurs when:

• A temporary database connectivity issue interrupted the request.
• A brief database failover or maintenance event is in progress.
• A network disruption between the service and its database occurred.

To resolve this error:

1. Wait at least a minute and try again — this error usually clears on its own.
2. If the error persists after several retries, check trust.salesforce.com for known incidents.
3. Avoid sending the same request repeatedly in quick succession — the database recovers faster when not under sustained load.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_DATABASE_UNAVAILABLE
• Whether the error is consistent or intermittent

A platform service the Model Builder depends on didn’t respond in time, so the request couldn’t complete. This is usually a temporary issue — wait a moment and try again.

This error occurs when:

• An upstream platform service was slow or under load when the request was made.
• A larger-than-usual workload caused the request to exceed the allowed time.
• A temporary network disruption delayed the upstream response.

To resolve this error:

1. Wait at least a minute and try again — this error usually clears on its own.
2. If the error persists after several retries, check trust.salesforce.com for known incidents.
3. If the request is unusually large, try breaking it into smaller pieces.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_DEPENDENCY_TIMEOUT
• Whether the error is consistent or intermittent

A platform service the Model Builder depends on is temporarily unavailable, so the request couldn’t complete. This is usually a temporary issue — wait a moment and try again.

This error occurs when:

• An upstream platform service is temporarily unavailable or undergoing maintenance.
• An upstream service is restarting or under high load.
• A temporary network disruption between platform services occurred.

To resolve this error:

1. Wait at least a minute and try again — this error usually clears on its own.
2. If the error persists after several retries, check trust.salesforce.com for known incidents.
3. Avoid sending the same request repeatedly in quick succession — waiting a few minutes between retries gives the dependency time to recover.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_DEPENDENCY_UNAVAILABLE
• Whether the error is consistent or intermittent

The task arguments in the request couldn’t be parsed. The values supplied aren’t in a form the service can process. Retrying without changes won’t help — the request needs to be corrected first.

This error occurs when:

• A required field inside taskArguments is missing.
• A field value has the wrong type — for example, a number where text is expected.
• The structure of taskArguments doesn’t match the schema for the operation being requested.
• The taskArguments payload is malformed.

To resolve this error:

1. Check the detail field in the error response — it identifies which part of the taskArguments failed parsing when available.
2. Compare your taskArguments against the documented schema for the operation you’re calling and correct any mismatches.
3. Resubmit the corrected request.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_DPC_PARAMS_INVALID
• A redacted copy of the taskArguments payload that was rejected

A training job with this name already exists in your org. Job names must be unique, so the request was rejected. Retrying with the same name won’t help.

This error occurs when:

• A previous request with the same job name already succeeded.
• An automated process is creating jobs with the same name and ran more than once.
• A retry of an earlier request that actually succeeded — the job was created but the response wasn’t received.

To resolve this error:

1. List your existing training jobs to confirm whether a job with the same name already exists. If it does, use that job instead of creating a new one.
2. If you need a new job, choose a unique name — consider appending a timestamp or unique identifier.
3. If your client is retrying after a timeout, check whether the original request succeeded before retrying.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_DUPLICATE_JOB_NAME
• The job name that was rejected

Your org has reached the maximum number of training jobs that can be queued or running at the same time. The new job was rejected. Wait for some in-progress jobs to finish, then try again.

This error occurs when:

• Too many jobs from your org are currently queued or in progress.
• An automated workflow is submitting jobs faster than they can complete.
• A backlog from a previous slowdown is still processing.

To resolve this error:

1. Wait for some of your in-progress jobs to finish, then try submitting the new job.
2. Check your job list to see what’s currently queued or running.
3. If you regularly hit this limit during normal use, contact Salesforce Customer Support to discuss your org’s quota.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_ENQUEUE_LIMIT_EXCEEDED
• A description of your typical job-submission volume

While processing your request, the Model Builder service called a platform dependency that rejected the data as invalid. Retrying without changes won’t help — the request needs to be corrected first.

This error occurs when:

• A value in your request becomes invalid only after the service applies additional context — for example, org-specific metadata — and the dependency then rejects the assembled request.
• A combination of values in your request is individually valid but jointly disallowed by the dependency.
• An identifier in your request doesn’t correspond to a record the dependency can see.

To resolve this error:

1. Check the detail field in the error response — it identifies the reason the dependency rejected the request when one is available.
2. Compare the values in your request against the API reference for the endpoint and correct anything that doesn’t satisfy the documented constraints.
3. Resubmit the corrected request.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_GRPC_INVALID_ARGUMENT
• A redacted copy of the request that triggered the error

This error code indicates an unexpected system error and not an issue with your configuration. If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_GRPC_UPSTREAM_ERROR
• Any context that might help with the investigation, such as when the error occurred, what you or your end user were doing at the time, and whether you’re able to reproduce it

This error code indicates an unexpected system error and not an issue with your configuration. If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_INTERNAL_ERROR
• Any context that might help with the investigation, such as when the error occurred, what you or your end user were doing at the time, and whether you’re able to reproduce it

The training job exists but is in a state where the requested operation isn’t allowed — for example, you can’t cancel a job that’s already completed. Retrying won’t help — check the job’s current status first.

This error occurs when:

• The job has already reached a terminal state — succeeded, failed, or cancelled — and the requested action no longer applies.
• The job is mid-transition and the requested action is only valid in a specific phase.
• Another process already moved the job out of the state this operation requires.

To resolve this error:

1. Fetch the job’s current status to see what state it’s in.
2. If the job is in a terminal state, no further action is needed — start a new job if you need to retry the work.
3. If the state isn’t what you expected, check whether another process acted on the job first.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_INVALID_JOB_STATE
• The job ID and the operation that was attempted

The request couldn’t be accepted because it failed validation. One or more required fields are missing, malformed, or outside their allowed values. Retrying without changes won’t help.

This error occurs when:

• A required field is missing from the request.
• A field value has the wrong type or format — for example, a number sent where text is expected.
• A field value is outside the allowed range or doesn’t match an allowed value.
• The request body is malformed or doesn’t match the expected structure.

To resolve this error:

1. Check the detail field in the error response — it identifies which part of the request failed validation.
2. Compare your request against the API reference for the endpoint and correct any missing, mismatched, or out-of-range values.
3. Resubmit the corrected request.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_INVALID_REQUEST
• A redacted copy of the request body that triggered the error

The training job has already been cancelled or has reached a terminal state where cancellation no longer applies. No further action will be taken on the job.

This error occurs when:

• The job was previously cancelled by you or another user with access to the same org.
• The job ran to completion before the cancellation request was received.
• An automated process issued a cancellation on a job that was already in a terminal state.

To resolve this error:

1. Fetch the job’s current status to confirm it’s already cancelled or in another terminal state.
2. If you intended to cancel a different job, verify the job ID in the cancellation request.
3. If you need to redo the work, start a new job — cancelled jobs can’t be resumed.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_JOB_ALREADY_CANCELLED
• The job ID that was the target of the cancellation

The training job referenced in the request couldn’t be found. Retrying with the same job ID won’t help.

This error occurs when:

• The job ID is mistyped or copied incompletely.
• The job belongs to a different org and isn’t visible to the current account.
• The job was deleted, archived, or removed by a retention policy.

To resolve this error:

1. Verify the job ID matches the value returned when the job was created or from a recent job list.
2. Confirm you’re using credentials for the org that owns the job.
3. List your org’s training jobs to confirm the job exists and check its current status.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_JOB_NOT_FOUND
• The job ID that was looked up

This error code indicates an unexpected system error and not an issue with your configuration. If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_METADATA_PARSE_ERROR
• Any context that might help with the investigation, such as when the error occurred, what you or your end user were doing at the time, and whether you’re able to reproduce it

The Model Builder service couldn’t complete the request because a call to the AI model infrastructure failed. This is a temporary platform-side issue — wait a moment and try again.

This error occurs when:

• The AI model infrastructure is temporarily unavailable or under load.
• A temporary network disruption occurred between the service and the AI model infrastructure.
• The AI model infrastructure returned an error that doesn’t map to a more specific failure.

To resolve this error:

1. Wait at least a minute and try again — this error usually clears on its own.
2. If the error persists after several retries, check trust.salesforce.com for known incidents.
3. Avoid sending the same request repeatedly in quick succession — waiting a few minutes between retries gives the dependency time to recover.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_MLLAKE_UPSTREAM_ERROR
• Whether the error is consistent or intermittent

This error code indicates an unexpected system error and not an issue with your configuration. If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_MTLS_CONFIG_MISSING
• Any context that might help with the investigation, such as when the error occurred, what you or your end user were doing at the time, and whether you’re able to reproduce it

The request was authenticated but the service couldn’t determine which org it applies to. Retrying with the same credential won’t help.

This error occurs when:

• The credential used to call the API isn’t associated with an org in this environment.
• A required org identifier is missing from the request.
• The account’s org association was removed or hasn’t been provisioned yet.

To resolve this error:

1. Confirm you’re using a credential that’s associated with an org in this environment.
2. Check with your Salesforce admin to verify your account is correctly provisioned for the org you’re trying to access.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_NO_TENANT_CONTEXT
• The type of credential being used — without exposing the credential value itself

The service couldn’t finalize a model artifact. A step in the artifact finalization process failed. This is usually a temporary issue — wait a moment and try again.

This error occurs when:

• A temporary failure occurred in the platform’s compute layer during one of the finalization steps.
• A temporary network disruption interrupted the finalization process.
• A previous finalization step left the artifact in an inconsistent state.

To resolve this error:

1. Wait at least a minute and try again — this error usually clears on its own.
2. If the error persists after several retries, contact Salesforce Customer Support — recovery may require platform-side cleanup.
3. Avoid sending the same request repeatedly in quick succession — waiting a few minutes between retries gives the platform time to recover.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_RAY_ARTIFACT_FAILURE
• The job or artifact identifier that was being finalized

The resource referenced in the request — a job, model setup, run, or related record — couldn’t be found. Retrying with the same identifier won’t help.

This error occurs when:

• The identifier is mistyped or copied incompletely.
• The resource belongs to a different org and isn’t visible to the current account.
• The resource was deleted, archived, or removed before this request was made.

To resolve this error:

1. Verify the identifier in the request matches the value returned when the resource was created or from a recent list.
2. Confirm you’re using credentials for the org that owns the resource.
3. List resources of that type to confirm whether the resource was ever created or has been removed.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_RESOURCE_NOT_FOUND
• The exact identifier that was looked up

A storage operation failed while processing the request. The service couldn’t read from or write to the artifact store. This is usually a temporary issue — wait a moment and try again.

This error occurs when:

• A temporary storage outage or throttling event affected the artifact store.
• A temporary network disruption occurred between the service and the artifact store.
• A storage location wasn’t ready or has moved.

To resolve this error:

1. Wait at least a minute and try again — most storage issues are temporary and clear on their own.
2. If the error persists after several retries, check trust.salesforce.com for known incidents.
3. Avoid sending the same request repeatedly in quick succession — storage services recover faster when not under sustained load.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_S3_ERROR
• Whether the error is consistent or intermittent

This error is logged internally by the Model Builder service’s background processing and isn’t normally surfaced in API responses. If a training job appears stalled and this code appears in a trace shared by Salesforce support, there’s nothing to do on your end — the service retries automatically on the next processing cycle.

This error occurs when:

• A temporary failure occurred while processing an internal background message.
• A temporary disruption in the internal message queue affected processing.
• A malformed message in the queue couldn’t be processed.

To resolve this error:

1. If a training job appears stuck, wait a few minutes — the service retries automatically.
2. If the job remains stuck after an extended period, contact Salesforce Customer Support.

If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_SQS_PROCESSING_ERROR
• The job ID that appears stalled, if known

This error code indicates an unexpected system error and not an issue with your configuration. If the issue persists, contact Salesforce Customer Support with the following information.

• The trace ID from the error response (instance field, format urn:trace:...)
• The error code: MBS_SWEEPER_FAILURE
• Any context that might help with the investigation, such as when the error occurred, what you or your end user were doing at the time, and whether you’re able to reproduce it

j

Get Started

Agentforce Examples

Agent Script

Agentforce DX

Agentforce APIs and SDKs

Agentforce Actions

Agentforce Service Errors

Language Detection Service Errors

Agent Service Reasoner Errors

Authentication Service Errors

Model Builder Errors

Interactive Data Science Service Errors

ML Lake Service Errors

Models and Prompts

Headless Agentforce Development