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

Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

developers.googleblog.com

Post reply on HN
Print view View original

41–50 of 151 posts

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#41

Post by simonw »

GaggiX wrote: post #23
simonw wrote: post #13

This price drop is significant. For For comparison, GPT-4o is currently $5/million input and $15/million output and Claude 3.5 Sonnet is $3/million input and $15/million output. Gemini 1.5 Pro was already the cheapest of the frontier models and now it's even cheaper.

GPT-4o is 2.5/10$. Unless you look at an old checkpoint. GPT-4o was the cheapest frontier model before.

I can’t see that price on https://openai.com/api/pricing/ - it’s listing $5/m input and $15/m output for GPT-4o right now.

No wait, correction: That’s confusing: it lists 4o first and then lists gpt-4o-2024-08-06 as $2.50/$10.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#42

Post by spiralk »

999900000999 wrote: post #28
diggan wrote: post #25

Earlier quoted context omitted.

Makes sense, as soon as your data leaves your computer, it's safe to assume it's no longer private, no matter what promises a service gives you. You want guaranteed private data that won't be used for anything? Keep it on your own computer.

Not sure why your getting down voted. Anything sent to an cloud hosted LLM is subject to be publicly released or used in training. Setting up a local LLM isn't that hard, although I'd probably air gap anything truly sensitive. I like ollama, but it wouldn't surprise me if it's phoning home.

This is not true. Both OpenAI and Google's LLM APIs have a policy of not using the data sent over them. Its no different than trusting Microsoft's or Google's cloud to store private data.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#44

Post by victor106 »

jzebedee wrote: post #34

As someone who actually had to build on Gemini, it was so indefensibly broken that I couldn't believe Google really went to production with it. Model performance changes from day to day and production is completely unstable as Google will randomly decide to tweak things like safety filtering with no notice. It's also just plain buggy, as the agent scaffolding on top of Gemini will randomly fail or break their own int…

Same experience here.

I had hopes of Google able to compete with Claude and OpenAI. But I don’t think that’s the case. Unless they come out with a product that’s 10x better in the next year or so I think they lost the AI race.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#46

Post by Workaccount2 »

summerlight wrote: post #12

Looks like they are more focused on the economical aspect of those large models? Like 90~95% performance of other frontier models at 50%~70% price.

They are going for large corporate customers. They are a brand name with deep pockets and a pretty risk-adverse model.

So even if Gemini sucks, they'll still win over execs being pushed to make a decision.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#47

Post by phren0logy »

spiralk wrote: post #42
999900000999 wrote: post #28

Earlier quoted context omitted.

Not sure why your getting down voted. Anything sent to an cloud hosted LLM is subject to be publicly released or used in training. Setting up a local LLM isn't that hard, although I'd probably air gap anything truly sensitive. I like ollama, but it wouldn't surprise me if it's phoning home.

This is not true. Both OpenAI and Google's LLM APIs have a policy of not using the data sent over them. Its no different than trusting Microsoft's or Google's cloud to store private data.

Can you link to documentation for Google's LLMs? I searched long and hard when Gemma 2 came out, and all of the LLM offerings seemed specifically exempted. I'd love to know if that has changed.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#48

Post by spiralk »

phren0logy wrote: post #47
spiralk wrote: post #42

Earlier quoted context omitted.

This is not true. Both OpenAI and Google's LLM APIs have a policy of not using the data sent over them. Its no different than trusting Microsoft's or Google's cloud to store private data.

Can you link to documentation for Google's LLMs? I searched long and hard when Gemma 2 came out, and all of the LLM offerings seemed specifically exempted. I'd love to know if that has changed.

https://ai.google.dev/gemini-api/terms this?

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#49

Post by kendallchuang »

spiralk wrote: post #32
kendallchuang wrote: post #24

Has anyone used Gemini Code Assist? I'm curious how it compares with Github Copilot and Cursor.

The Aider leaderboards seem like a good practical test of coding usefulness: https://aider.chat/docs/leaderboards/ . I haven't tried Cursor personally but I am finding Aider with Sonnet more useful that Github Copilot and its nice to be able to pick any model API. Eventually even a local model may be viable. This new Gemini model does not rank very high unfortunately.

Thanks for the link. That's unfortunate, though perhaps the benchmarks will be updated after this latest Gemini release. Cursor with Sonnet is great, I'll have to give Aider a try as well.

Re: Two new Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

#50

Post by Workaccount2 »

xendipity wrote: post #36
sgt101 wrote: post #29

Earlier quoted context omitted.

We might see that with the inference ASICs later this year I guess?

Ooh, what are these ASICs you're talking about? My understanding was that we'll see AMD/Nvidia gpus continue to be pushed and very competitive as well as have new system architectures like cerebras or grok. I haven't heard about new compute platforms framed as ASICs.

Cerebras has ridiculously large LLM ASICs that can hit crazy speeds. You can try it with llama 8B and 70B:

https://inference.cerebras.ai/

It's pretty fast, but my understanding is that it is still too expensive even accounting for the speed-up.

Post reply on HN