thanks for the great talk . quick question - what are your go to models for on prem? for example, something like gpt-oss (without fine tuning) would be best open source, that balances resources vs. capabilities e.g. tool calling? @wide-doctor-37830 @wooden-umbrella-84098

Thanks for joining Elvin! @chilly-france-44439 do you have good recommendations here?

It depends how large a model you can deploy, qwen-3-coder-480b and glm-4.6 are pretty good. Gpt oss also kinda works