Dear Beshar,
I hope this message finds you well.
As we all know, Codex offers two powerful capabilities: "goal" and context compression. These features significantly improve long-context handling and task-oriented workflows.
Would it be possible to add similar functionality to the DeepSeek API? Specifically, I would like to request:
Goal support — allowing users to set a persistent high-level objective that guides the model across multiple turns.
Context compression — a mechanism to summarize or compress conversation history automatically, reducing token usage while preserving key information.
These additions would make the DeepSeek API much more competitive for complex, multi-step tasks.
Thank you for considering this request. I would be happy to provide more details or test early implementations if helpful.
Best regards,
hapusama
Dear Beshar,
I hope this message finds you well.
As we all know, Codex offers two powerful capabilities: "goal" and context compression. These features significantly improve long-context handling and task-oriented workflows.
Would it be possible to add similar functionality to the DeepSeek API? Specifically, I would like to request:
Goal support — allowing users to set a persistent high-level objective that guides the model across multiple turns.
Context compression — a mechanism to summarize or compress conversation history automatically, reducing token usage while preserving key information.
These additions would make the DeepSeek API much more competitive for complex, multi-step tasks.
Thank you for considering this request. I would be happy to provide more details or test early implementations if helpful.
Best regards,
hapusama