[Bug]: Getting "ValidationException: The provided model identifier is invalid." for Bedrock provide

[nisranjan]

@augustintsang

Environment (please complete the following):

• OS: [e.g. Ubuntu 22.04] GNU/Linux 5.15.164.4-microsoft-standard-WSL2
• kubectl-ai version (run kubectl-ai version): [e.g. 0.3.0] 0.3.0
• LLM provider: [e.g. gemini, openai, grok...] Amazon Bedrock
• LLM model: [e.g. gemini-2.5-pro] Amazon Titan Text Express V1

Describe the bug
I am trying to solve for issues in bedrock and coming across an error. The specific error is

E0903 10:33:59.626219 1 conversation.go:472] "error sending streaming LLM response" err="bedrock stream error: operation error Bedrock Runtime: ConverseStream, https response error StatusCode: 400, RequestID: a9464e5c-4679-4f86-bf6d-46cf872c8bd3, ValidationException: The provided model identifier is invalid."

I am using LLM Model Amazon Titan Text Express V1 (since Claude was not available) and I have turned off Systems prompt in code as per this document but still getting the same error.

I have tried to use other models also, for e.g. Amazon Nova Pro. for also but here also I get the same error. Actually in this case when I try to go from AWS CLI, I get a different error

**C:\Users\DELL\OneDrive\Documents\kubectl-ai>aws bedrock-runtime converse --region ap-south-1 --model-id amazon.nova-pro-v1:0 --messages file://messages.json --profile nranjan

An error occurred (ValidationException) when calling the Converse operation: Invocation of model ID amazon.nova-pro-v1:0 with on-demand throughput isn’t supported. Retry your request with the ID or ARN of an inference profile that contains this model.**

I could not use Anthropic Claude Sonnet models, because I made some error while getting approval to use them from Amazon Marketplace and now I am not able to get that approval - at least I don't have it now.

To Reproduce
Steps to reproduce the behavior:

0. Setup environment variables, the following
   AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY & AWS_REGION

1. Go to my kubectl-ai project and build an image using
   docker build -f images/kubectl-ai/Dockerfile -t kubectl-ai:latest .

2. Run command '...'
   docker run --rm -it -v ~/.kube:/root/.kube -v ~/home/ubuntu/.aws:/root/.aws -e AWS_ACCESS_KEY_ID -e AWS_SECRET_ACCESS_KEY -e AWS_REGION=us-east-1 -e BEDROCK
   _MODEL="amazon.titan-text-express-v1" kubectl-ai:v0.1 --llm-provider=bedrock "test"

3. See errorExpected behavior
   E0903 10:33:37.477675 1 conversation.go:472] "error sending streaming LLM response" err="bedrock stream error: operation error Bedrock Runtime: ConverseStream, https response error StatusCode: 400, RequestID: ce65503d-1fdd-4d5b-b821-168c4f51207e, ValidationException: The provided model identifier is invalid."

Additional context
Add any other context or logs here.

[droot]

@augustintsang can you pl. help since you are the most familiar with the bedrock.

[droot]

Can you try with the latest release v0.0.24 ?

[nisranjan]

@droot @augustintsang

I have closed these in PR #629
Pls review