[ENHANCEMENT] Turn on prompt caching for supported Cerebras model `zai-glm-4.7`

[jahanson]

Problem (one or two sentences)

Users are unable to take advantage of the newly supported prompt caching feature by Cerebras' model zai-glm-4.7.
The provider's page clearly states prompt caching is supported: https://inference-docs.cerebras.ai/models/zai-glm-47

Context (who is affected and when)

Whenever the model zai-glm-4.7 is selected in the Cerebras provider, it is set as disabled.

Desired behavior (conceptual, not technical)

Enable prompt caching by default.

Constraints / preferences (optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear context and impact

Roo Code Task Links (optional)

No response

Acceptance criteria (optional)

No response

Proposed approach (optional)

Set the model zai-glm-4.7's prompt caching to true.

Roo-Code/packages/types/src/providers/cerebras.ts

Line 24 in a33117a

supportsPromptCache: false,

Trade-offs / risks (optional)

No response

[roomote]

I have analyzed this issue and here is my proposed implementation plan:

Change: In packages/types/src/providers/cerebras.ts, update the zai-glm-4.7 model configuration to enable prompt caching:

"zai-glm-4.7": {
    maxTokens: 16384,
    contextWindow: 131072,
    supportsImages: false,
-   supportsPromptCache: false,
+   supportsPromptCache: true,
    supportsNativeTools: true,
    defaultToolProtocol: "native",
    inputPrice: 0,
    outputPrice: 0,
    description: "Highly capable general-purpose model on Cerebras..."
},

This aligns with the Cerebras documentation confirming that zai-glm-4.7 supports prompt caching.

Please mention @roomote in a comment saying that you approve this plan, and I will proceed with the implementation.

[hannesrudolph]

@roomote go for it

[roomote]

Thanks for the approval! I'll implement the change to enable prompt caching for the zai-glm-4.7 model now. I'll open a draft PR shortly.

[roomote]

Opened PR #10670 with my attempt at addressing this. Feedback welcome!