SEZ9 opened a new issue, #11547: URL: https://github.com/apache/seatunnel/issues/11547
### Search before asking - [X] I had searched in the [feature](https://github.com/apache/seatunnel/issues?q=is%3Aissue+label%3A%22Feature%22) and found no similar feature requirement. ### Description Some OpenAI-family models on AWS Bedrock (e.g. `openai.gpt-5.6-terra`, `openai.gpt-5.6-sol`) are not listed in the foundation-model catalog and reject all standard Bedrock runtime APIs (`Converse`, `InvokeModel`, `ChatCompletions`). Per the [AWS model card](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-terra.html) they are served exclusively through the OpenAI **Responses API** on the dedicated `bedrock-mantle` endpoint (`https://bedrock-mantle.{region}.api.aws/openai/v1`), authenticated with a Bedrock bearer token instead of SigV4. The SeaTunnel CLI's existing providers (`bedrock` via Converse, `openai` via Chat Completions) cannot reach these models: - `bedrock` provider: `ValidationException: The provided model identifier is invalid` - `openai` provider pointed at the mantle base URL: `The model 'openai.gpt-5.6-terra' does not support the '/v1/chat/completions' API` ### Usage Scenario Users whose LLM access is standardized on AWS Bedrock (credentials, billing, governance) want to use these OpenAI models with the SeaTunnel CLI for natural-language config generation, without needing a separate OpenAI account. Proposed: a `bedrock-mantle` provider that - talks to the Responses API on the mantle endpoint, - derives a short-term bearer token from the caller's AWS credentials (`aws-bedrock-token-generator`) with automatic refresh, - converts between the CLI's internal message format and Responses API items (including function-call round-trips and streaming events), - omits `temperature` (unsupported by these models). Usage: `AI_PROVIDER=bedrock-mantle OPENAI_MODEL=openai.gpt-5.6-terra seatunnel "..."` ### Related issues #11508 (the CLI benchmark work that surfaced this gap) ### Are you willing to submit a PR? - [X] Yes I am willing to submit a PR! ### Code of Conduct - [X] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
