refactor: improve retrieve model (#1036)

- check the model type while retrieve model
- select chat/reranker model even if it is missed in client models
- find predefined-models for openai-compatible client with startsWith
- remove client::ApiType

1 files changed, 3 insertions, 3 deletions

diff --git a/src/client/vertexai.rs b/src/client/vertexai.rs
index 07ff2a4..7b73164 100644
--- a/src/client/vertexai.rs
+++ b/src/client/vertexai.rs
@@ -45,7 +45,7 @@ impl Client for VertexAIClient {
         let model = self.model();
         let model_category = ModelCategory::from_str(model.name())?;
         let request_data = prepare_chat_completions(self, data, &model_category)?;
-        let builder = self.request_builder(client, request_data, ApiType::ChatCompletions);
+        let builder = self.request_builder(client, request_data);
         match model_category {
             ModelCategory::Gemini => gemini_chat_completions(builder, model).await,
             ModelCategory::Claude => claude_chat_completions(builder, model).await,
@@ -63,7 +63,7 @@ impl Client for VertexAIClient {
         let model = self.model();
         let model_category = ModelCategory::from_str(model.name())?;
         let request_data = prepare_chat_completions(self, data, &model_category)?;
-        let builder = self.request_builder(client, request_data, ApiType::ChatCompletions);
+        let builder = self.request_builder(client, request_data);
         match model_category {
             ModelCategory::Gemini => {
                 gemini_chat_completions_streaming(builder, handler, model).await
@@ -84,7 +84,7 @@ impl Client for VertexAIClient {
     ) -> Result<Vec<Vec<f32>>> {
         prepare_gcloud_access_token(client, self.name(), &self.config.adc_file).await?;
         let request_data = prepare_embeddings(self, data)?;
-        let builder = self.request_builder(client, request_data, ApiType::Embeddings);
+        let builder = self.request_builder(client, request_data);
         embeddings(builder, self.model()).await
     }
 }