update model

.env

@@ -3,5 +3,5 @@ ALLOW_ORIGINS=*
 GMAIL_POLL_SECONDS=30
 VERTEX_PROJECT_ID=flow-pilot-493104
 VERTEX_LOCATION=us-central1
-VERTEX_MODEL=gemini-2.5-pro
+VERTEX_MODEL=gemini-2.5-flash
 GOOGLE_APPLICATION_CREDENTIALS=

.env.example

@@ -3,5 +3,5 @@ ALLOW_ORIGINS=*
 GMAIL_POLL_SECONDS=30
 VERTEX_PROJECT_ID=your-gcp-project-id
 VERTEX_LOCATION=us-central1
-VERTEX_MODEL=gemini-2.5-pro
+VERTEX_MODEL=gemini-2.5-flash
 GOOGLE_APPLICATION_CREDENTIALS=

.github/workflows/deploy.yml

@@ -58,4 +58,4 @@ jobs:
           image: ${{ env.REGION }}-docker.pkg.dev/${{ env.PROJECT_ID }}/${{ env.REPOSITORY }}/${{ env.IMAGE }}:${{ github.sha }}
           flags: >-
             --allow-unauthenticated
-            --set-env-vars=AI_PROVIDER=vertex_ai,VERTEX_PROJECT_ID=${{ env.PROJECT_ID }},VERTEX_LOCATION=${{ env.REGION }},VERTEX_MODEL=gemini-2.5-pro
+            --set-env-vars=AI_PROVIDER=vertex_ai,VERTEX_PROJECT_ID=${{ env.PROJECT_ID }},VERTEX_LOCATION=${{ env.REGION }},VERTEX_MODEL=gemini-2.5-flash

README.md

@@ -74,5 +74,6 @@ The workflow builds the Docker image, pushes it to Artifact Registry, and deploy
 - Fill in your real Vertex values in `.env`, especially `VERTEX_PROJECT_ID` and `VERTEX_LOCATION`.
 - If you authenticate with `gcloud auth application-default login`, leave `GOOGLE_APPLICATION_CREDENTIALS` blank.
 - Only set `GOOGLE_APPLICATION_CREDENTIALS` when you have a real service account JSON path available.
+- The default model is `gemini-2.5-flash` for better latency and cost during iteration.
 - The backend now attempts live Vertex AI calls through the `google-genai` SDK when credentials are configured.
 - If the SDK is missing or `VERTEX_PROJECT_ID` is still a placeholder, FlowPilot falls back to local deterministic mock logic so tests and scaffolding still work.

backend/config.py

@@ -16,7 +16,7 @@ class Settings(BaseModel):
     gmail_poll_seconds: int = int(os.getenv("GMAIL_POLL_SECONDS", "30"))
     vertex_project_id: str = os.getenv("VERTEX_PROJECT_ID", "your-gcp-project-id")
     vertex_location: str = os.getenv("VERTEX_LOCATION", "us-central1")
-    vertex_model: str = os.getenv("VERTEX_MODEL", "gemini-2.5-pro")
+    vertex_model: str = os.getenv("VERTEX_MODEL", "gemini-2.5-flash")
     google_application_credentials: str = os.getenv(
         "GOOGLE_APPLICATION_CREDENTIALS",
         "",