Increased Composer Context Window

Feature request for product/service

AI Models

Describe the request

Composer needs to have the ability to significantly increase its context window size. I am working on a problem set, that literally takes the entire context for it to figure out what it needs to do, and right as it gets smart enough, it summarizes context, and its dead.

The only way I can get through these problemsets (which are constant and frequent) is to switch to GPT 5.4 and use the 1M context window. I can solve most of these problems I’m working with about 250-500k context. Well after putting in about $1k for the month, I’ve been cut off and have had to go back to using Composer. It’s made every inch gained, take easily 5-7x longer.

Most problem sets do not require a massive context window but a project like the one I am working on, has so many complex interconnections, and debugging required to get a handle on “what’s actually happening” that this, is the case.

For “context”, its a transpiler/compiler project, that is taking wicked real world wild code complexity and converting it. I’ve done ALOT to reduce the workflow down, to exactly how one should go about it, but the type probing, debugging code emission, takes alot of context. Once it has it, boom, dead. It’ll throw out all the processes, and start modifying all over the place.

Operating System (if it applies)

Linux

Hi @leadership Thanks for the feedback! We are aware of the ask, and we appreciate you weighing in on this. Any model updates will be announced loudly here in the forum.

Thanks @kevinn .

@kevinn Is there a roadmap or ticket to track progress on this request? 200K tokens context window is too small compared to modern models (OpenAi, Anthropic, DeepSeek, etc.). Thank you.