AI deployed model not working. No support nothing. this is In Azure AI hub

Question

AI deployed model not working. No support nothing. this is In Azure AI hub

Dhanasimmaraja S 0

https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com/swagger.json I deployed this In Azure AI hub but its not working. tested all endpoints but no luck. - ✅ Successfully fetched swagger from: https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com/swagger.json

INFO - 📋 Analyzing Swagger Endpoints

INFO - Found 17 paths in swagger documentation

INFO - 📍 POST /v2/online_anomaly_detection: Timegen

INFO - 📍 POST /v2/cross_validation: Timegen

INFO - 📍 POST /v2/anomaly_detection: Timegen

INFO - 📍 POST /v2/historic_forecast: Timegen

INFO - 📍 POST /v2/forecast: Timegen

INFO - 📍 POST /cross_validation_multi_series: Timegen

INFO - 📍 POST /anomaly_detection_multi_series: Timegen

INFO - 📍 POST /historic_forecast_multi_series: Timegen

INFO - 📍 POST /forecast_multi_series: Timegen

INFO - 📍 POST /historic_forecast: Timegen

INFO - 📍 POST /forecast: Timegen

INFO - 📍 POST /validate_api_key: Utilities

INFO - 📍 POST /validate_token: Utilities

INFO - 📍 POST /model_input_size: Utilities

INFO - 📍 POST /model_params: Utilities

INFO - 📍 GET /info: Info

INFO - 📍 GET /listRoutes: Get All Routes

INFO -

🚀 Testing All Discovered Endpoints

INFO - Base URL: https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com

INFO - Total endpoints to test: 17

INFO -

==================================================

INFO - Testing: POST /v2/online_anomaly_detection

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported, please contact support"}

INFO -

==================================================

INFO - Testing: POST /v2/cross_validation

INFO - Summary: Timegen

INFO - Using basic test payload

INFO - ❌ HTTP 422: {"detail":[{"type":"missing","loc":["body","series"],"msg":"Field required"},{"type":"missing","loc":["body","freq"],"msg":"Field required"},{"type":"missing","loc":["body","n_windows"],"msg":"Field r

INFO -

==================================================

INFO - Testing: POST /v2/anomaly_detection

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported, please contact support"}

INFO -

==================================================

INFO - Testing: POST /v2/historic_forecast

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported, please contact support"}

INFO -

==================================================

INFO - Testing: POST /v2/forecast

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported, please contact support"}

INFO -

==================================================

INFO - Testing: POST /cross_validation_multi_series

INFO - Summary: Timegen

INFO - Using basic test payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /anomaly_detection_multi_series

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /historic_forecast_multi_series

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /forecast_multi_series

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /historic_forecast

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /forecast

INFO - Summary: Timegen

INFO - Using time series payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: POST /validate_api_key

INFO - Summary: Utilities

INFO - Using basic test payload

INFO - ❌ HTTP 405: {"detail":"Method Not Allowed"}

INFO -

==================================================

INFO - Testing: POST /validate_token

INFO - Summary: Utilities

INFO - Using basic test payload

INFO - ✅ SUCCESS: 200 (1.51s)

INFO - 📊 Response size: 34 bytes

INFO - 📄 Response: {'detail': 'Happy Forecasting! :)'}

INFO -

==================================================

INFO - Testing: POST /model_input_size

INFO - Summary: Utilities

INFO - Using basic test payload

INFO - ❌ HTTP 404: {"detail":"Not Found"}

INFO -

==================================================

INFO - Testing: POST /model_params

INFO - Summary: Utilities

INFO - Using basic test payload

INFO - ❌ HTTP 422: {"detail":"Model 'timegpt-1' is not supported."}

INFO -

==================================================

INFO - Testing: GET /info

INFO - Summary: Info

INFO - Using basic test payload

INFO - ✅ SUCCESS: 200 (1.11s)

INFO - 📊 Response size: 193 bytes

INFO - 📄 Response: {'model_name': 'nixtla-timegen1', 'model_type': 'forecasting', 'model_provider_name': 'Nixtla', 'served_model_name': 'nixtla-timegen1', 'served_model_type': 'forecasting'}

INFO -

==================================================

INFO - Testing: GET /listRoutes

INFO - Summary: Get All Routes

INFO - Using basic test payload

INFO - ✅ SUCCESS: 200 (1.40s)

INFO - 📊 Response size: 1011 bytes

INFO - 📄 Response: [{'path': '/swagger.json', 'name': 'openapi'}, {'path': '/docs', 'name': 'swagger_ui_html'}, {'path': '/docs/oauth2-redirect', 'name': 'swagger_ui_redirect'}, {'path': '/redoc', 'name': 'redoc_html'}, {'path': '/v2/online_anomaly_detection', 'name': 'timegen'}, {'path': '/v2/cross_validation', 'name

INFO -

======================================================================

INFO - COMPREHENSIVE ENDPOINT ANALYSIS

INFO - ======================================================================

INFO - Total endpoints tested: 17

INFO - Successful: 3

INFO - Failed: 14

INFO -

✅ WORKING ENDPOINTS:

INFO -

🏥 HEALTH/STATUS ENDPOINTS:

INFO - ✅ GET https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com/info (1.11s)

INFO -

🔧 OTHER WORKING ENDPOINTS:

INFO - ✅ POST https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com/validate_token (1.51s)

INFO - ✅ GET https://TimeGEN-1-zeolh.eastus2.models.ai.azure.com/listRoutes (1.40s)

SRILAKSHMI C 10,885 Reputation points Microsoft External Staff Moderator

2025-11-07T07:03:31.0433333+00:00

Hi Dhanasimmaraja S,

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank you!

1 answer

Your answer

SRILAKSHMI C 10,885 Reputation points Microsoft External Staff Moderator

2025-11-07T07:03:31.0433333+00:00

Hi Dhanasimmaraja S,

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank you!

Answer 1

Hello Dhanasimmaraja S,

Welcome to Microsoft Q&A and Thank you for sharing the detailed information and logs.

I understand that your deployed TimeGEN model in Azure AI Hub is returning errors such as “Model ‘timegpt-1’ is not supported” across multiple endpoints, even though the Swagger documentation loads successfully.

This behavior usually indicates a backend configuration or model availability issue, but there are a few checks you can perform to rule out setup and payload-related causes:

1. Model Availability

Please verify whether the timegpt-1 model is supported in the East US 2 region. Some models are not yet available in all regions and deploying them can result in unsupported model errors. You can check the Azure model regional availability documentation to confirm.

2. Check Your Payloads

Some of the 422 errors also indicate missing fields (e.g., series, freq, n_windows) that are required for specific endpoints like /v2/cross_validation and /v2/forecast. Ensure your payloads include all required parameters as specified in the Swagger documentation before sending requests.

3. Subscription Header

Confirm that your API requests include the proper subscription header for authentication. Example:

{ "apim-subscription-id": "<your-subscription-id>" }

Missing this header can cause certain endpoints to reject requests even if the deployment looks active.

4. Inspect Model Configuration

Check your AI Hub deployment settings to ensure the TimeGEN model is configured properly:

Open your deployed model in Azure AI Hub.

Verify the Model Provider field it should show Nixtla - TimeGEN1 or a supported version.

If it still lists timegpt-1, try redeploying the model using the latest available TimeGEN version from the catalog.

Please refer this Troubleshoot the multivariate API,

I Hope this helps. Do let me know if you have any further queries.

Thank you!

Dhanasimmaraja S 0 Reputation points

2025-11-18T08:39:47.51+00:00

Region is supported tried deployed again, I see the model is unavailable in MS side
SRILAKSHMI C 10,885 Reputation points Microsoft External Staff Moderator

2025-11-21T17:04:07.7233333+00:00

Hi Dhanasimmaraja S,

Could you please share the exact commands you are using? This will help us verify whether the syntax is supported.

Thank you!

Share via

AI deployed model not working. No support nothing. this is In Azure AI hub

1 answer

Your answer