i for one am very glad to have a true non-thinking qwen3 with absolutely bonkers benchmarks and none of that /no_think <think></think> business

@TheAhmadOsman i don’t need it to beat opus or kimi. i just want it to be like V3-0324 but smaller

@TheAhmadOsman it’s on i like it. very impressed from initial checks. catches itself after minor stumbles, seems to know what it knows/doesn’t, rare world knowledge is decent, refuses over-hard questions, pleasant assistant persona, not too sloppy, tool-calls well

@TheAhmadOsman will run some custom benches in a bit but could def see it replacing 4.1/V3 as default API model i hope they give 30B-A3B the same treatment…