Live data from Hacker News

Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

blog.google

611–616 of 616 posts

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#611

Earlier quoted context omitted.

I find Simon's work informative and entertaining; the last thing he can be accused of is low effort. The Pelicans are just a bit of fun icing on top.

Sending a one sentence prompt to an LLM and posting it to hackernews constantly isn’t low effort? Today I learnt something new

if that's all he did sure, he makes an interesting blog nearly every day jealous or what?

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#612
post #360

Earlier quoted context omitted.

You need to compare cost per task buddy boy. Cost per token doesn't tell you much when you don't know how many tokens a model will use to accomplish a task

>boy Seriously?

buddy man

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#613

Earlier quoted context omitted.

Could it be that they have to serve their models to billions of users?

> Could it be that they have to serve their models to billions of users? And how is that different from their competitors exactly?

Their competitors aren't able to serve models as good as Gemini 3.6 Flash to billions of users

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#614
post #321

Earlier quoted context omitted.

It's literally the meme of the MS org chart pointing guns at each other. The GCP team wants their slice, the other team wants some otjer slice, and so on. Everyone wants some crap for their promotion package. It's no wonder Meta has shit the bed even worse. It's also why Google still releases actually decent, useful models despite the product being such a hilarious mess. A lot of the time Gemini models have actually…

Exactly my experience. I'm building an AI document-extraction platform, so I had to benchmark a bunch of models — on the cost:quality:latency:adherence picture, flash wins hands down for structured extraction. (Caveat: I've only tested the three US labs and Mistral.). So like u said, totally viable in prod for a relatively static tool. Didn't build the tool suite with gemini, but if you use service mode it currently…

[dead]

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#615

It's scary relying on Google's models. I have a very price sensitive workload that used to run on flash 2.5 lite - it's deprecated now. The replacement 3.1 flash lite is a lot more expensive, but now also has a sunset date. 3.5 flash lite is even more expensive. So the price is rising and you have no choice but to keep paying more and more.

I'm running price-sensitive data extraction workloads on flash 2.5 and its still the king when it comes to accuracy + cost, all the gemini 3 variants perform a bit worse and cost a lot more. Low-key freaking out, ngl

you should try out scaledown, it's 29x cheaper and 9% more accurate than gemini https://scaledown.ai/benchmarks/scaledown-vs-gemini

Re: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

#616

Earlier quoted context omitted.

This drove me bonkers. You can enable play music (etc) in Android Auto for workspace accounts by enabling apps in Gemini. From memory (looking at the settings now, not 100% sure of the magic steps required), but go to admin.google.com, go to 'generative ai', 'gemini app', and 'apps settings', then turn on 'other Google apps'. This lets you play music (and other things) in Android Auto.

Ah, that could be it. I saw that "Other Google apps" and didn't think that would mean Spotify, but I guess its Android Auto or Google Assistant stuff. I'll give that a try, thanks for the tip.

As an update, I enabled that setting and I'm still not able to have Gemini actually be useful for calling people or changing playlists in my car.
Post reply on HN