Update gemini 2 flash thinking model name check

[Gasser-Aly]

a Small change to update the check for gemini 2 flash thinking model to use contains instead of == to allow specifying a region if needed with OpenAI compatible proxies.

Ex: usage:

client<llm> GEMINI_V2_FLASH_THINKING {
  provider "openai-generic"
  retry_policy Exponential
  options {
    model "googlevertexai-us-central1:gemini-2.0-flash-thinking-exp-1219"
    api_key env.MY_TOKEN
    base_url "https://my-proxy.ai/v1"
  }
}

---

Important

Update content_part() in googleai_client.rs to use contains for model name check, allowing region-specific names.

• Behavior:
  • Update content_part() in googleai_client.rs to use contains instead of == for model name check, allowing region-specific model names.
• Misc:
  • No other changes or refactoring in the code.

^{This description was created by for 9f19ef8. It will automatically update as commits are pushed.}

[vercel]

Someone is attempting to deploy a commit to the Gloo Team on Vercel.

A member of the Team first needs to authorize it.

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to 9f19ef8 in 12 seconds

More details

• Looked at 17 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. engine/baml-runtime/src/internal/llm_client/primitive/google/googleai_client.rs:470

• Draft comment:
  Consider adding a comment to content_part to explain the use of contains for model name matching, as it allows for more flexible specification, including regions.
• Reason this comment was not posted:
  Confidence changes required: 33%
  The change from == to contains is appropriate for the described use case, allowing for more flexible model name matching. However, the function content_part should be documented to reflect this change and its implications.

Workflow ID: wflow_mfIb1xzRh1u3tiG9

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[sxlijin]

Thanks! We'll get this merged asap