The past little while has been filled with new LLM releases. I am curious to understand what you find to work best for your particular workflows and tasks.
Qwen 3.8 27b-mlx running locally on Apple M4 48 GB. I only use local LLMs, mostly for getting technical suggestions and code snippet samples. Qwen 3.8 has been the first local model I've tried that has consistently given usable and impressive results.
Astra on Extra High. Knowledge work, the right answer is much more important than a fast or cheap answer. If I were to switch to Anthropic, Opus 5.5 or Fable 5.1 on high or better.
My preferred model at the moment is GLM 5.3 Coding. I think it's as good as the latest closed-weight models at a fraction of the cost, and we are being capped now at my place of employment so if I use it I don't run out of quota by the end of the first week of each month.
in codex - luna high for 80% of everything. This keeps me within the $100 a month plan. Then every so often Sol 6 high or even Astro xhigh but rarely. Also a lot of chat on xhigh or astro because that's billed seperate so I can have chat do a code review of what luna high just did and paste it to luna. Luna is always so embarrassed and agrees with the review.
gpt 5.6 sol. assuming the prose is similar with the gpt 6 series, if it was available at work that's probably what i'd be using instead.
i'm well aware of opus and fable's superior engineering qualities for writing code, however i find that as a swe who needs to understand every engineering decision that is made and is held accountable for their actions for safety critical code, understanding trumps speed every time.
and the gpt model series is far and above easier for me to understand the outputs of than claude, even with the prose improvements in fable 5.1 and opus 5.5.
i'm well aware of opus and fable's superior engineering qualities for writing code, however i find that as a swe who needs to understand every engineering decision that is made and is held accountable for their actions for safety critical code, understanding trumps speed every time.
and the gpt model series is far and above easier for me to understand the outputs of than claude, even with the prose improvements in fable 5.1 and opus 5.5.