Skip to content

Update cerebras.ts #9024

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Nov 5, 2025
Merged

Update cerebras.ts #9024

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6260be2
Update cerebras.ts
sebastiand-cerebras Nov 4, 2025
9edc4d2
reduced to 16k tokens
sebastiand-cerebras Nov 4, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions packages/types/src/providers/cerebras.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,13 +7,13 @@ export const cerebrasDefaultModelId: CerebrasModelId = "gpt-oss-120b"

export const cerebrasModels = {
"zai-glm-4.6": {
maxTokens: 16_384,
contextWindow: 128000,
maxTokens: 40960,
Copy link
Collaborator

@mrubens mrubens Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hey @sebastiand-cerebras, I'm up for whatever you all think is best but personally I think reserving that many tokens for output is a little much for a 131k context window.

Copy link
Author

@sebastiand-cerebras sebastiand-cerebras Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi Matt,

Thank you so much for your message! We decided to go for this value across our external messaging but I see that it's rather a theoretical value. Would you want me to reduce it in the case of roocode? Wouldn't be an issue!

Btw, I am a big fan :)

Copy link
Collaborator

@mrubens mrubens Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you! Yeah personally I've found 16k to be enough for coding, and then you reserve ~115k for task history before condensing etc.

Copy link
Author

@sebastiand-cerebras sebastiand-cerebras Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I updated the maxTokens. What do you think?

Copy link
Collaborator

@mrubens mrubens Nov 5, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good, thanks!

contextWindow: 131072,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
description: "Highly intelligent general-purpose model with ~2000 tokens/s",
description: "Highly intelligent general purpose model with up to 1,000 tokens/s",
},
"qwen-3-coder-480b-free": {
maxTokens: 40000,
Expand Down