I tried it myself it with argon, gemini 3.8 flash, beam, and others models and found that models like beam perform bad and also just don't agree to do it at all so its mostly like a thing on flagship models like gpt astra and claude opus and fable like models
You attempted on closed weights models? That might be why