feat(together-ai): update model YAMLs [bot]

[models-bot]

Auto-generated by poc-agent for provider together-ai.

---

Note

Low Risk
Catalog-only metadata changes with no application logic; incorrect costs or flags could misroute clients or billing estimates.

Overview
Bot sync updates five Together AI model YAMLs so routing, pricing, and capability flags match current provider listings.

pearl-ai/gemma-4-31b-it gets higher per-token costs, new function_calling / json_output features, and thinking: true. Qwen3-235B-A22B-Instruct-2507-FP8 adds structured_output. vidu/vidu-q3-turbo now accepts image inputs (R2V-style) and drops its sources block. alibaba/happyhorse-1.0-r2v and allenai/Molmo-7B-D-0924 are marked status: active, with a pricing comment tweak on happyhorse.

^{Reviewed by Cursor Bugbot for commit dcd30d2. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test setup failed for together-ai

The test job aborted before any tests ran. Error:

Failed to apply provider-account/together-ai 'test-v2-together-ai': {"statusCode":500,"message":"Error applying the manifest of type provider-account"}

This is usually a transient infra issue (catalogue build, TrueFoundry API, GitHub archive). Try /test-models again, or check the job logs.