feat(models): declare reasoning efforts for glm-5.2 on bytedance

[AmineAce]

Summary

Declares reasoningEfforts for glm-5.2 on providerId: "bytedance" in packages/models/src/models/zai.ts.

Changes

• glm-5.2 (bytedance, glm-5-2-260617) → ["none", "minimal", "low", "medium", "high", "xhigh", "max"]

Source

Volcengine Chat API reference (Chinese docs). glm-5-2-260617 is the one model explicitly named on Volcengine's allowlist for the full tier range:

• none and xhigh: restricted to glm-5-2-260617 only.
• max: allowed for glm-5-2-260617, deepseek-v4-pro-260425, deepseek-v4-flash-260425.
• minimal/low/medium/high: documented without model restriction.

Combined, this gives the full 7-tier ladder for this specific model/host — directly confirmed, not inferred.

Not changed

• Other glm-5.2 hosts checked but excluded: canopywave and granite (no documented reasoning-control parameter at all), embercloud (the reasoning.effort parameter exists in their docs, but no accepted values are published — nothing concrete to cite).

Testing

• pnpm format — clean, only zai.ts changed (array wrapped across multiple lines by the formatter).

Summary by CodeRabbit

• New Features
  • Added support for selecting reasoning effort levels—from none to maximum—when using the GLM-5.2 model through the ByteDance provider.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 9b1eea56-0cfa-4a19-8e76-c98c04f53b15

📥 Commits

Reviewing files that changed from the base of the PR and between 45baf8d and 87caf9c.

📒 Files selected for processing (1)

• packages/models/src/models/zai.ts

---

Walkthrough

The ByteDance provider mapping for glm-5.2 now includes an explicit list of supported reasoning effort levels: none, minimal, low, medium, high, xhigh, and max.

Changes

GLM-5.2 reasoning effort configuration

Layer / File(s)	Summary
Add reasoning effort allowlist packages/models/src/models/zai.ts	The ByteDance configuration for glm-5.2 declares supported reasoning effort levels from none through max.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Possibly related PRs

• theopenco/llmgateway#3012: Uses provider reasoningEfforts configuration when setting reasoning effort in requests.
• theopenco/llmgateway#3072: Adds the same per-provider reasoningEfforts configuration to another model mapping.
• theopenco/llmgateway#3131: Adds explicit reasoning effort arrays for another ByteDance model mapping.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: adding reasoning effort declarations for glm-5.2 on the Bytedance provider.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}