In Pi I am seeing some kind of default context size limit 128k for models:
gemini:models/gemini-3.5-flash-lite (from gemini configured in otari)zen:space-bunny-freeopenrouter:nvidia/nemotron-3-ultra-550b-a55b:free
and probably more. Is this issue of pi-otari or otari? The same models accessed via directly configured providers in Pi seem to have proper context size limit.
In Pi I am seeing some kind of default context size limit 128k for models:
gemini:models/gemini-3.5-flash-lite(from gemini configured in otari)zen:space-bunny-freeopenrouter:nvidia/nemotron-3-ultra-550b-a55b:freeand probably more. Is this issue of pi-otari or otari? The same models accessed via directly configured providers in Pi seem to have proper context size limit.