Same impression here. If this is Google's strategy we'll have to migrate to Luna, because costs are already high for our workload on 3.8 Flash (which in itself was already a 40% increase compared to 3.7 flash due to being more verbose and thinking more) and this now doubles cost once again.
Re:Gemini 4 Hydrogen
by ComputerGuru on Thursday November 12, 2026 @01:28PM
(#50471887)
Yeah, and they rug-pulled 3.7 very quickly, silently redirecting to 3.8 Flash. Raised our prices and broke some workflows.
Re:Gemini 4 Hydrogen
by alexgoodhart on Thursday November 12, 2026 @04:14PM
(#50474374)
Yeah they fuckin did
Re:Gemini 4 Hydrogen
by amarcheschi on Thursday November 12, 2026 @01:27PM
(#50471857)
Isn't something open weight cheaper (and probably better as well)?
Re:Gemini 4 Hydrogen
by ComputerGuru on Thursday November 12, 2026 @01:30PM
(#50471910)
Depends on your needs. Gemini Flash has best bang for your buck on visual tasks and best multilingual support/intelligence for non-Western/non-East-Asian languages; better than paid or OSS alternatives.
Re:Gemini 4 Hydrogen
by amarcheschi on Thursday November 12, 2026 @01:36PM
(#50472008)
Yes I gave for granted textual tasks but they're not the only ones. If I remember correctly, mistral large should be good as well on image tasks while being cheap(er). I never had to do anything with less common languages tho
all of google's actions in AI lately seem to have the same explanation: they'd rather you not use their models through the API, they have better things to do with that compute than rent it to you. the rate limits on argon and carbon, the pricing, the general difficulty of access.
this is probably the new flash-lite equivalent, and they probably could serve it for flash-lite pricing. but they don't want to, because then they'd have less capacity for serving it in their own products.
In the Gemini app it's replaced Flash. Lite is still there.
Re:Gemini 4 Hydrogen
by irthomasthomas on Thursday November 12, 2026 @01:36PM
(#50472003)
But that's an old model, and if they replace it they have no smaller element to use for it's name. I guess it would have to be Hydrogen Light.
I suppose they could name it proton or hydron.
Re:Gemini 4 Hydrogen
by agildehaus on Thursday November 12, 2026 @01:44PM
(#50472126)
Quark/Neutrino/Axion/Tachyon.
Re:Gemini 4 Hydrogen
by iamdelirium on Thursday November 12, 2026 @01:36PM
(#50472000)
The fact that there is still a flash lite dropdown in Gemini points to them keeping the flash lite model family.
Re:Gemini 4 Hydrogen
by irthomasthomas on Thursday November 12, 2026 @01:37PM
(#50472021)
So all models will be named after elements except Flash Light? Or maybe the next version will be named Hydrogen Light?
Gemini Electron?
Isn't Flash-lite or related model the one used for AI summary in Google Search?
> considerably higher than what anyone is paying today for 3.8 Flash or 3.5 Flash Lite or the deprecated 3.0/2.x Gemini models
It's cheaper than the last Pro models.
Re:Gemini 4 Hydrogen
by wasabi991011 on Thursday November 12, 2026 @03:21PM
(#50473627)
According to Artificial Analysis, it's much cheaper per task than 3.8 flash, 3$ for H xhigh vs 6-7$ for 3.8 medium-high. I don't understand how though, since AA says it has more output tokens per task.
But it is priced the same as the non-discounted 3.8 Flash, considerably higher than what anyone is paying today for 3.8 Flash or 3.5 Flash Lite or the deprecated 3.0/2.x Gemini models. How does this work? Is Google abandoning the “fast and cheap” tier? That seems like a mistake since plenty of workflows are absolutely price inelastic, meaning when the 3.8 Flash discount ends at EOY, lots of current enterprise Gemini customers would have to take their business to another shop??