Skip to content
This repository was archived by the owner on May 15, 2026. It is now read-only.

feat: add glm-5-turbo model to Z.ai provider - #12022

Draft
ghost wants to merge 1 commit into
mainfrom
feature/add-glm-5-turbo-model
Draft

feat: add glm-5-turbo model to Z.ai provider#12022
ghost wants to merge 1 commit into
mainfrom
feature/add-glm-5-turbo-model

Conversation

@ghost

@ghost ghost commented Mar 27, 2026

Copy link
Copy Markdown

This PR attempts to address Issue #12018.

Changes

Adds glm-5-turbo to both international and mainland Z.ai model lists in packages/types/src/providers/zai.ts.

Model specs (international):

  • Context window: 200,000 tokens
  • Max output tokens: 128,000
  • Input price: $1.2/M tokens
  • Cached input price: $0.24/M tokens
  • Output price: $4.0/M tokens
  • Prompt caching: supported

Mainland pricing

Estimated proportionally from the existing glm-5 international/mainland price ratio.

Testing

  • All 33 existing Z.ai tests pass
  • Lint and type checks pass across the entire monorepo

Feedback and guidance are welcome.

Interactively review PR in Roo Code Cloud

Add glm-5-turbo to both international and mainland Z.ai model lists
with 200k context window, 128k max output tokens, and prompt caching.

International pricing: $1.2/M input, $0.24/M cached, $4.0/M output
Mainland pricing estimated proportionally from glm-5 ratios.

Closes #12018
@akfrolov

Copy link
Copy Markdown

Maybe support for GLM-5-Turbo and GLM-5.1 (#12024) could be implemented together?

@MobCode100

Copy link
Copy Markdown

glm-5-turbo should support reasoning. I think these lines should be added:

		supportsReasoningEffort: ["disable", "medium"],
		reasoningEffort: "medium",
		preserveReasoning: true,

Tests also be added for this model in zai.spec.ts.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants