Tested Omen Alpha High on my hardest 3D benchmark with architectural constraint, and it’s not good.
It ignored architecture rules, threw immediate TypeScript errors, and performed worse than Qwen 3.8 Flash (120b) and GLM 5.3 Flash. feels like a 50-120b checkpoint from GLM
DeepSeek V4 Flash isn’t giving me the results I’m looking for.
Going to try the MiaAI_lab Qwen 3.8 Flash for one spark and let you know what I think.
I gave 4 "Flash" models the same job: redact the PII in an HR letter using an agentic mask tool.
One of them cost 630× more than another. And it wasn't 630× better. Results 🧵 🥇 GLM 5.3 Flash $0.005, 15 tool calls Clean sweep. Overshot a bit (also redacted employee ID +
I have some MTP dense benchmarks Qwen 3.8 27B Q4 MTP vs Qwen 3.6 35B Q4 MTP....
Poor 3.8 27b lol
You have reached the end of the archive
All of qwen38