[Docs] Document image input support for DeepSeek V4 models

[zoomote]

Opened on behalf of Toray Altas. View the task or mention @zoomote for follow-up asks.

What changed

Adds an "Image Input" note to the DeepSeek provider page stating that deepseek-v4-pro and deepseek-v4-flash support image (vision) inputs, while the legacy deepseek-chat and deepseek-reasoner aliases do not.

Why this change was made

The paired Zoo Code fix enables image support for the DeepSeek V4 models; the docs should reflect the corrected capability so users know they can attach images with these models.

Impact

Documentation-only change. pnpm run check-types and the Docusaurus build (pre-push hook) pass.

Related PRs

• Code fix: [Fix] Images rejected as unsupported when using DeepSeek V4 Pro or V4 Flash Zoo-Code#969

Linked work items

Closes Zoo-Code-Org/Zoo-Code#964

[zoomote]

1 issue outstanding. See task

• docs/providers/deepseek.md:50 — The new "Image Input" note documents a capability that only becomes true once the paired code PR [Fix] Images rejected as unsupported when using DeepSeek V4 Pro or V4 Flash Zoo-Code#969 merges (currently open/draft; on main both V4 models still have supportsImages: false). Please hold this docs PR until the code fix ships, or the docs will claim image support that the released extension still rejects with unsupported_model.

_{Reviewed 4f91a5c}