Hey, as of right now nothing has changed. There’s still no separate setting to manually set the context window for a custom OpenAI-compatible model, and there’s also no auto-detection of the real limit from the provider. I can’t share an ETA.
Quick note for your case: for custom models that aren’t in our catalog, the default is 1M. If you’re seeing less (like 200K), it’s most likely because the model name matches a catalog entry that’s capped lower. Can you tell me which model you’re using and what exact name you’re registering it under? I’ll take a look at what’s happening. As a workaround, you can try registering it under a name that isn’t in the catalog, then the default 1M should apply.
The most helpful thing right now is to add your case (model, provider’s real limit, and that you specifically need 1M) to this feature request: Unlock Full Context Window with Own API Keys This helps with prioritization.