> ## Documentation Index > Fetch the complete documentation index at: https://docs.synapsai.cloud/llms.txt > Use this file to discover all available pages before exploring further. # Troubleshooting > Common API and deployment errors, what they mean, and how to fix them If a model deployment failed, check the model **Logs** tab for full error details before taking action. Below are common error names, messages, probable causes, and fixes. If you are still stuck, [contact support](/support/contacting-support) with the model ID, timestamp, and error message. *** ## Authentication and permissions ### `InvalidAPIKeyError` **Message:** `Invalid or restricted API key.`\ **Cause:** API key is missing, expired, revoked, or restricted.\ **Fix:** Confirm the key, check status in the dashboard, generate a new key if needed, and send `Authorization: Bearer `. ### `PermissionError` **Message:** `Permission error:
`\ **Cause:** Account or key lacks permission for the resource.\ **Fix:** Verify role/permissions and ensure the key belongs to a user with access. ### `InvalidHFTokenError` **Message:** `Invalid Hugging Face token, please check the token and try again.`\ **Cause:** HF token is invalid or expired.\ **Fix:** Re-enter a valid HF token in model settings and redeploy. *** ## Billing, quotas, and spend control ### `InsufficientCreditsError` **Message:** `Insufficient credits. Please add more credits to your account.`\ **Cause:** Account balance is too low.\ **Fix:** Add credits on the [Billing](https://platform.synapsai.cloud/billing) page or reduce deployment cost. See [Billing](/manage/billing). ### `MaxTokensError` **Message:** `The maximum allowed number of tokens for this completion has been reached.`\ **Cause:** Request exceeded token limits.\ **Fix:** Reduce prompt size, lower `max_tokens`, or increase context in deployment settings. *** ## Model lifecycle and deployment ### `ModelLoadError` **Message:** `Model load error, please see model logs for more details`\ **Cause:** Inference server failed to load model files.\ **Fix:** Inspect logs. Verify file layout, VRAM, quantization, and retry deployment. ### `ModelNotFoundError` **Message:** `Model not found`\ **Cause:** Wrong or deleted model ID.\ **Fix:** Confirm model ID from the dashboard. ### `ModelNotReadyError` **Message:** `Model not ready`\ **Cause:** Model is still being provisioned, building, or an instance is still loading.\ **Fix:** Retry after a moment. If the model is **Building**, wait for deployment to finish. If errors persist, check logs for stalled provisioning. ### `ModelBuildingError` **Message:** `It is not possible to query a model that is building.`\ **Cause:** Model is in the build phase.\ **Fix:** Wait and monitor build logs. ### `DeploymentFailedError` **Message:** `It is not possible to query a failed model.` or `Deployment failed, please check the deployment logs for more details.`\ **Cause:** Deployment failed.\ **Fix:** Inspect deployment logs. Contact [support](/support/contacting-support) with logs and model ID if needed. ### `FailedStressTestError` **Message:** `Failed to stress test the model.`\ **Cause:** Model did not pass stability/throughput tests.\ **Fix:** Check logs for memory/timeout spikes; try smaller batch size or different hardware. ### `InvalidQuantizationError` **Message:** `Invalid quantization.`\ **Cause:** Unsupported or corrupted quantization.\ **Fix:** Redeploy with a supported precision/quantization. *** ## Input, preprocessing, and templates ### `PreprocessingError` **Message:** `There was an error with the preprocessing of the inputs.`\ **Cause:** Preprocessors failed on audio, video, or image inputs.\ **Fix:** Validate and re-encode inputs; test with text-only input to isolate the issue. ### `URLContentLoadError` **Message:** `Failed to the content on url in the model's inputs.`\ **Cause:** URL unreachable or unsupported content type.\ **Fix:** Verify the URL is publicly accessible or upload content directly. ### `InvalidParameterError` **Message:** `Invalid parameter:
`\ **Cause:** Missing or malformed request parameter.\ **Fix:** Validate request schema against the [API reference](/api-reference/introduction) and [Supported tasks](/resources/tasks). ### `InvalidMessageError` **Message:** `Invalid message:
`\ **Cause:** Message content violates chat endpoint schema.\ **Fix:** Confirm conversation/message structure matches the expected JSON schema. ### `InvalidConversationError` **Message:** `The structure of the conversation is invalid.`\ **Cause:** Conversation object has missing fields or invalid order.\ **Fix:** Reformat the conversation payload. ### `ChatTemplateError` **Message:** `There was an error with the chat template when applying it to the inputs.`\ **Cause:** Chat template placeholders do not match inputs.\ **Fix:** Update the chat template in model settings or adjust input keys. ### `NoChatTemplateError` **Message:** `No chat template found for this model`\ **Cause:** Model expects a `chat_template` for chat endpoints.\ **Fix:** Provide `chat_template` in the request or configure a default in model settings. *** ## Endpoints and capabilities ### `EndpointNotSupportedError` **Message:** `The endpoint is not supported for this model.`\ **Cause:** The model does not expose that endpoint.\ **Fix:** Use a supported endpoint from [Supported tasks](/resources/tasks) or redeploy with the required pipeline. ### `PrecisionRequiredError` **Message:** `No default precision found for this model.`\ **Cause:** Model requires explicit precision/quantization.\ **Fix:** Set precision in deployment settings and redeploy. *** ## Hugging Face and repository ### `InvalidHFModelError` **Message:** `Invalid Hugging Face model name`\ **Cause:** Incorrect or missing HF repo name.\ **Fix:** Verify repo name and access. ### `RevisionNotFoundError` **Message:** `Provided revision for this model was not found`\ **Cause:** Branch/tag/commit does not exist.\ **Fix:** Use a valid revision or omit to use default. ### `GatedRepoError` **Message:** `This model is gated and requires a valid token`\ **Cause:** HF repo access is restricted.\ **Fix:** Provide a valid HF token with repo permissions. ### `DisabledRepoError` **Message:** `This model on Hugging Face is disabled`\ **Cause:** HF repo disabled by owner or platform.\ **Fix:** Re-enable repo or use a different model. ### `NoSafetensorsError` **Message:** `No safetensors file found on this model repository`\ **Cause:** Repo lacks Safetensors weights.\ **Fix:** Follow [Convert models to Safetensors](/guides/convert-safetensors) and redeploy. ### `RemoteCodeError` **Message:** `Remote code error, we do not support remote code for models.`\ **Cause:** Model requires executing remote code at load time.\ **Fix:** Use a model without remote code or contact [support](/support/contacting-support). ### `HfHubHTTPError` / `HuggingFaceError` **Cause:** Hugging Face Hub returned an error while fetching assets.\ **Fix:** Check HF status, token validity, and repo configuration in logs. *** ## Runtime and internal ### `InternalError` **Message:** `Internal error:
`\ **Cause:** Unexpected server-side failure.\ **Fix:** Check logs and report to support with timestamps. ### `APIError` **Cause:** Generic API failure.\ **Fix:** Inspect request/response details and server logs. ### `RopeScalingError` **Message:** `The backend failed to apply RoPE scaling to the model.`\ **Cause:** RoPE scaling failed for large context or incompatible weights.\ **Fix:** Reduce context size or redeploy with different config. *** ## When to contact support If logs and the fixes above do not resolve the issue, contact [support](/support/contacting-support) with: * Model ID and deployment ID * Timestamp of the failing request * Error name and message from the UI or API response * Steps to reproduce Check [Status and uptime](https://status.synapsai.cloud) for platform-wide incidents before opening a ticket.