feat(server, templates): add Veo 3.1 model support

[buildingvibes]

Description

Add support for Google's new Veo 3.1 video generation models across the Echo router, SDK, and video template. This resolves #595 by adding both veo-3.1-generate-preview and veo-3.1-fast-generate-preview as supported models and making them the default in the video template.

Changes

Router (packages/app/server)

• Added veo-3.1-fast-generate-preview and veo-3.1-generate-preview to the VEO3_MODELS list in VertexAIProvider.ts
• Updated the test client (veo3-client.ts) to use the new 3.1 model

SDK (packages/sdk/ts)

• Added Veo 3.1 models to Vertex AI video model definitions with pricing ($0.40/sec with audio for standard, $0.15/sec for fast — same rate as Veo 3 for 720p/1080p)
• Added Veo 3.1 models to Gemini video model definitions (veo-3.1-generate-001, veo-3.1-fast-generate-001)
• Updated TypeScript type unions to include the new model IDs

Template (templates/next-video-template)

• Added Veo 3.1 Fast and Veo 3.1 to the model dropdown in the video generator UI
• Set Veo 3.1 Fast as the new default model
• Updated the API validation to accept the new model strings
• Updated the vertex handler type signature
• Updated page comments to reflect the new model support

Testing

• Verified all type definitions are consistent across SDK, server, and template
• Validated that existing Veo 3.0 models are preserved for backward compatibility
• Pricing data sourced from Google's official Gemini API pricing page

Fixes #595

[vercel]

@buildingvibes is attempting to deploy a commit to the Merit Systems Team on Vercel.

A member of the Team first needs to authorize it.

[buildingvibes]

Closing this PR as the CI requires team authorization for fork deploys which I cannot resolve from my side. Happy to reopen if a team member can authorize the Vercel deployment. Thank you!