Support doubleword flex tier

https://docs.doubleword.ai/inference-api/async-inference

The issue references the DoubleWord AI async inference API. Look for existing provider integrations in the codebase, likely under a directory like `src/providers` or `src/backends`. Examine how other async inference services are implemented to understand the required traits and configuration. The goal is to add a new provider module that can make requests to the DoubleWord API and handle responses.

Do mô hình lập chỉ mục viết ra từ nội dung của issue.