Validate AI Model

Used to validate an AI model's API configuration by testing the connection to the model provider. This ensures that the model credentials and configuration are correct before using it in production.

API Endpoint

Property	Value
Request Method	POST
Request URL	https://api.seliseblocks.com/models/{model_id}/validate

Request

Request Example

curl -X POST 'https://api.seliseblocks.com/models/model_123/validate?project_key=YOUR_PROJECT_KEY' \
  -H 'accept: application/json'

Path Parameters

Field	Type	Required	Description
model_id	string	Yes	The unique identifier of the model to validate.

Query Parameters

Field	Type	Required	Description
project_key	string	No	The project key for your project.

Request Headers

Field	Type	Required	Description
accept	string	Yes	Accepted response format. Use application/json

tip

When to validate your AI model:

• After initial model creation to verify configuration
• Before deploying to production environment
• After updating API keys or credentials
• When troubleshooting connection issues
• After changing provider settings or base URLs
• Periodically to ensure continued connectivity

The validation process:

• Tests authentication with the provider's API
• Verifies the model exists and is accessible
• Checks API quota and rate limits
• Validates custom headers and parameters
• Confirms base URL connectivity
• Returns detailed error messages for debugging

Response

Success Response (200 OK)

Returns an object containing the validation results.

{
  "is_valid": true,
  "provider": "openai",
  "model_name": "gpt-4-turbo-preview",
  "connection_status": "successful",
  "api_version": "v1",
  "timestamp": "2025-01-12T10:30:00Z"
}

Response Fields

Field	Type	Description
is_valid	boolean	Indicates whether the model configuration is valid.
provider	string	The AI model provider name.
model_name	string	The specific model name being validated.
connection_status	string	Status of the connection test (e.g., "successful", "failed").
api_version	string	The API version being used.
timestamp	string	ISO 8601 timestamp of the validation.

note

Response Object Structure

The actual response structure may vary based on the provider. The response returns a flexible object (additionalProp1: {}) that can contain provider-specific validation details such as:

• Model availability status
• API endpoint health
• Quota information
• Supported features
• Provider-specific metadata

Error Response (422 Unprocessable Entity)

Returns validation error details when the model configuration is invalid or the connection test fails.

{
  "detail": [
    {
      "loc": [
        "path",
        "model_id"
      ],
      "msg": "Model not found or invalid model ID",
      "type": "value_error.notfound"
    }
  ]
}

Error Response Fields

Field	Type	Description
detail	array	Array of validation error objects.
loc	array	Location of the error in the request (e.g., path, query).
msg	string	Human-readable error message.
type	string	Error type identifier.

Common Validation Failures

Failure Type	Description	Resolution
Invalid API Key	API key is incorrect, expired, or revoked	Update model with valid API key
Model Not Found	Model ID doesn't exist in the system	Verify model_id is correct
Connection Timeout	Cannot reach provider's API endpoint	Check network connectivity and base_url
Authentication Failed	Provider rejected the credentials	Verify API key permissions and account status
Quota Exceeded	API usage limits reached	Upgrade plan or wait for quota reset
Invalid Configuration	Missing required provider-specific fields	Check Azure deployment_name or other required fields
Provider Unavailable	Model provider service is down	Wait and retry, or check provider status page

Error Codes

Status Code	Description	Response Type
200	OK - Model configuration is valid	Success
400	Bad Request - Invalid request format	Bad Request
401	Unauthorized - Invalid API credentials	Unauthorized
404	Not Found - Model ID doesn't exist	Not Found
422	Validation Error - Model configuration invalid	Unprocessable Entity
503	Service Unavailable - Provider API unavailable	Service Unavailable

warning

Important Considerations

• Validation tests make actual API calls to the provider, which may count against your quota
• Some providers charge for validation requests
• Validation does not guarantee all model features will work, only basic connectivity
• Failed validation prevents the model from being used in AI agents
• Always validate after updating model credentials or configuration
• Keep validation frequency reasonable to avoid unnecessary API costs