Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Please don't discontinue Gemini 2.5 Flash

discuss.ai.google.dev

Post reply on HN
Print view View original

61–70 of 93 posts

Re: Please don't discontinue Gemini 2.5 Flash

#61

Post by bjord »

dacox wrote: post #35
hrpnk wrote: post #9

I love how there is a "Please do not discontinue gemini-2.0-flash[-lite], 2.5 is NOT an equivalent" from Feb 20th. Getting too attached to models is a smell.

the 1.5 and 2.0 flash models were absolute beasts. They were very cheap, and _very_ fast. We contemplated moving some of our fine tuned workloads to them because we would have gotten very substantial total latency reductions for our workloads. However, they are aggressively deprecating them (OpenAI is as well), and replacing with newer models. These newer models are all reasoning models, and importantly, only bear th…

have you experimented at all with the deepseek flash models?

Re: Please don't discontinue Gemini 2.5 Flash

#62

Post by exabrial »

bjord wrote: post #60
exabrial wrote: post #58

I don't know how many times people will need to learn this: Do not use Google in Production.

in this case, the other frontier shops are all doing the same thing, so maybe the advice here should be "don't use closed weight models in production"?

Google has a history of pulling the rug on their paying customers and offering 0 support when it is convenient for them. You have no recourse. There are a billion cloud providers to choose from, this is not the first time this has been on the first page of HN.

Re: Please don't discontinue Gemini 2.5 Flash

#64

Post by bjord »

exabrial wrote: post #62
bjord wrote: post #60

Earlier quoted context omitted.

in this case, the other frontier shops are all doing the same thing, so maybe the advice here should be "don't use closed weight models in production"?

Google has a history of pulling the rug on their paying customers and offering 0 support when it is convenient for them. You have no recourse. There are a billion cloud providers to choose from, this is not the first time this has been on the first page of HN.

I'm aware. regardless, when it comes to models, the advice to "not use Google in Production" falls short.

Re: Please don't discontinue Gemini 2.5 Flash

#65

Post by recursive »

avaer wrote: post #18

Why not a "stop killing AI" movement? If a company deploys a paid AI model and makes people depend on it, they need to dump the weights at EOL.

If you paid a one-time fee for an offline model, and then you were revoked access to it, that would apply.

If you are paying an ongoing subscription for a service, I'd advise you not to rely on it too much or keep a list of alternatives.

Re: Please don't discontinue Gemini 2.5 Flash

#66

Post by imhoguy »

samuelknight wrote: post #23
hrpnk wrote: post #9

I love how there is a "Please do not discontinue gemini-2.0-flash[-lite], 2.5 is NOT an equivalent" from Feb 20th. Getting too attached to models is a smell.

It's not a smell. Why should these developers rebuild a core piece of their stack every few months. Switching out a model requires a new round of testing and validation when we should be able to rely on a piece of software the behave the same way since the last time we touched it.

This is how development looks like for many years now, constant rewrite on the horizon. I think LLM development hype surpassed Blockchain and JS frameworks craze of decade ago.

Re: Please don't discontinue Gemini 2.5 Flash

#67

Post by recursive »

derektank wrote: post #45
wat10000 wrote: post #22

Earlier quoted context omitted.

There are plenty of open weights models available already. If the ability to keep running the same model is important to you, then choose one of those.

Presumably, a “Stop Killing AI” movement, mirroring the Stop Killing Games movement, would require a provider that revokes access a previously available model to make it open weights at the time of death.

These are not analogous. If you paid for an offline model, and you were somehow revoked access because it was phoning home, that would be closer.

If you were paying for an ongoing subscription for a service, that would be something different.

Re: Please don't discontinue Gemini 2.5 Flash

#69

Post by mips_avatar »

llmslave wrote: post #68
mips_avatar wrote: post #2

It's such a good model for the price, for a lot of tasks it outperforms gpt5 at 3x the speed and 1/5 the price. The price jump from 2.5->3->3.5 has been so high.

i have found google models outperforming other models in actual agentic workflows

I find that Gemini flash 2.5 performs about as well as Claude sonnet for non coding agentic flows except it’s actually fast enough

Re: Please don't discontinue Gemini 2.5 Flash

#70

Post by imhoguy »

dacox wrote: post #35
hrpnk wrote: post #9

I love how there is a "Please do not discontinue gemini-2.0-flash[-lite], 2.5 is NOT an equivalent" from Feb 20th. Getting too attached to models is a smell.

the 1.5 and 2.0 flash models were absolute beasts. They were very cheap, and _very_ fast. We contemplated moving some of our fine tuned workloads to them because we would have gotten very substantial total latency reductions for our workloads. However, they are aggressively deprecating them (OpenAI is as well), and replacing with newer models. These newer models are all reasoning models, and importantly, only bear th…

I test workloads with multiple closed and at least one open model now. Good to have a backup on 503s or credits run out.
Post reply on HN