[Workers AI] Document model-specific reasoning controls

[KastanDay]

Stacked on #33022. Until that PR merges, GitHub includes its GLM-5.2 changes in this PR's diff. The layer-only comparison shows this PR's four-file change.

Why

Gemma 4 and DeepSeek V4 do not use the standard reasoning effort scale. Their model pages omit the accepted values and defaults, so users cannot tell which controls work or how compatibility aliases behave.

What changed

• List the accepted reasoning_effort values in the model schemas and Model Info cards.
• Document Gemma 4's binary reasoning control and its accepted aliases.
• Document DeepSeek V4's off-by-default behavior, supported tiers, aliases, and reasoning continuation behavior.

Logic diff: +21 / -0 lines

The count includes #33022 while both PRs target production.

Validation

• pnpm check
• pnpm lint
• pnpm format:check
• pnpm build (8,902 pages)
• pnpm test (124 tests)

[cloudflare-docs-bot]

Review

⚠️ 1 warning, 💡 1 suggestion found in commit 6b60469.

👉 Fix in your agent 👈

Fix the following review findings in PR #33025 (https://github.com/cloudflare/cloudflare-docs/pull/33025).

Before making changes, review each finding and present a brief summary table:
- For each finding, state whether you agree, disagree, or need clarification
- If you disagree (e.g. the fix requires disproportionate effort for minimal benefit,
  or the finding is factually incorrect), explain why
- If you need clarification before deciding, ask those questions
- Then share your plan for which issues to tackle and in what order

After triaging, follow this order:
1. Post a comment on this PR for any findings you are skipping, with the finding ID and your reasoning.
2. Then commit the fixes for the legitimate findings.

The comment must come before the commit — the bot reads PR comments when a new
push triggers a review, so skip comments posted after the push will be missed.

---

## Code Review

### Warnings (1)

#### CR-67b38c238025 · Request parameters duplicated in output schema
- **File:** `src/content/workers-ai-models/deepseek-v4-pro-0813.json` line 2414
- **Issue:** The added `reasoning_effort` and `chat_template_kwargs` blocks appear again at much deeper indentation around lines 2414-2436 and 3685-3707, which places them inside the output schema rather than under the two input variants.
- **Fix:** Remove `reasoning_effort` and `chat_template_kwargs` from the output schemas; they are request parameters and should only be documented under `schema.input` (Prompt and Messages).

### Suggestions (1)

#### CR-d4a0e3d5c2ed · Incomplete reasoning_effort mapping description
- **File:** `src/content/workers-ai-models/deepseek-v4-pro-0813.json` line 316
- **Issue:** The `reasoning_effort` enum includes `low`, but the description only documents mappings for `none`, `minimal`, `medium`, `auto`, and `xhigh`.
- **Fix:** Extend the description to state how `low` is handled (e.g. `low maps to low`) so every supported enum value is documented.

Code Review

This code review is in beta and may not always be helpful — use your judgment.

Warnings (1)

File

Issue

workers-ai-models/deepseek-v4-pro-0813.json line 2414

Request parameters duplicated in output schema — The added reasoning_effort and chat_template_kwargs blocks appear again at much deeper indentation around lines 2414-2436 and 3685-3707, which places them inside the output schema rather than under the two input variants. Fix: Remove reasoning_effort and chat_template_kwargs from the output schemas; they are request parameters and should only be documented under schema.input (Prompt and Messages).

Suggestions (1)

File	Issue
workers-ai-models/deepseek-v4-pro-0813.json line 316	Incomplete reasoning_effort mapping description — The reasoning_effort enum includes low, but the description only documents mappings for none, minimal, medium, auto, and xhigh. Fix: Extend the description to state how low is handled (e.g. low maps to low) so every supported enum value is documented.

Conventions

_{No convention issues found.}

Style Guide Review

_{No style-guide issues found.}

Commands

Only codeowners can run commands. Post a comment with the command to trigger it.

Command	Description
/review	Runs a review now. Incremental if a prior review exists, full if not.
/full-review	Re-reviews the entire PR diff from scratch, ignoring incremental history. Useful after a rebase, when you want a fresh review, or if the bot gets out of sync and reports issues that no longer exist.
/ignore-review-limit	Permanently lifts the 2-review automatic limit for this PR. Future pushes will trigger reviews as normal.
/disable-auto-review	Stops automatic reviews from triggering on future pushes to this PR. Codeowners can still run /review or /full-review manually.
/rebase	Rebases the PR branch against production. On conflict, attempts to resolve automatically using AI. Stops with an explanation if confidence is not high enough.

[KastanDay]

Superseded by #33541