Grok 4.1 models being deprecated from xAI API without any close alternative with cheap pricing ...

@techdevnotes Sounds like the non reasoning flavors of 4.3 are coming very soon

@techdevnotes Uhhhhh…. @xai please keep some of the cheap models.

@techdevnotes It was too cheap! But nowadays you can just use cheaper OSS models on @OpenRouter that have similar preformance

Behind every AI breakthrough is infrastructure you never think about. Supermicro infrastructure solutions combine optimized cabling, routing, and port mapping to reduce time-to-online, while scalable water and dry cooling towers support thermal management and expansion.

@techdevnotes There will be new models in the coming weeks.

@techdevnotes Grok 4.3 is such a dud i'ts super slow you can't set a resoning effort please xAI talk to actual developers

@techdevnotes Biggest reason to build on top of open source ones .

@techdevnotes There should be default options, other wise there will always be developers in loop for these changes

Wow the trap xAI set: They deprecated their fast non-reasoning models and the only replacement non-reasoning model is still in beta. Meanwhile their “flagship” 4.3 is reasoning-locked with a TTFT that’s voice-incompatible. So if you need fast voice and want Grok, you’re forced onto a beta at 10x the cost. Wow!

@techdevnotes I loved the model for fast chat interaction or for quick web and X retrieval all for a pretty good price. Literally integrated it with all of my search tools for that exact reason. Such a shame it’s going to be discontinued.

@techdevnotes This sucks. While lots of people focus on “frontier” and “state of the art,” but I have a lot of tasks for which Grok 4.1 Fast is “sufficient.” It does what it’s supposed to, reliably, fast and cheap. Now everything is slower and costs 500% more (but is slightly better.)

@techdevnotes Maybe the non-reasoning variant will be like 4.1

Bring ideas to life with Grok Imagine. Generate stunning AI images instantly.

@techdevnotes Ugh 4.1 fast reasoning was great for many small tasks

@techdevnotes deprecating a tier without keeping a cheap equivalent is the part that hurts integrators most. forced upgrade path = forced cost increase

@techdevnotes One week notice? Wow.

@techdevnotes Quand il est pas en panne alors...😂

@techdevnotes there are other alternatives just as cheap...just not grok

@techdevnotes $1.25M in, $2.50M out is pretty cheap….

@techdevnotes just cycling UP!

@techdevnotes That’s frustrating. It feels like good options are getting harder to find.

@techdevnotes Yeah, this is end of cheap AI. Now what?

@techdevnotes tf