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

Gemini 3.5 Flash

blog.google

Post reply on HN
Print view View original

421–430 of 692 posts

Re: Gemini 3.5 Flash

#421

Post by p1esk »

stared wrote: post #375
irthomasthomas wrote: post #161

Earlier quoted context omitted.

This is a perfect illustration of something I noticed with llm progress. Ask them to improve an svg like this, and it never fixes the missing crossbar or disconnected limbs, it just adds more stuff. In this example they have obviously improved greatly, and it contains a ridiculous amount of detail, but they still to get the basic shape of the frame wrong. It's weird. And the pattern shows up everywhere, try it with a…

To a certain extent, it feels like a Sonnet 3.7 moment. Slightly overeager - you ask for a button color change, you see layout changes, new package dependencies, and the README rewritten from scratch - and not necessarily correctly. When I ask for a pelican on a bike, I want the Platonic ideal of a pelican on a bike, not a vision of an alternative reality in which pelicans created bikes. Though, thinking about it aga…

What is “Sonnet 3.7 moment”?

Re: Gemini 3.5 Flash

#422

Post by kristopolous »

esafak wrote: post #372
kristopolous wrote: post #262

I have a tool to track these I've built Relatively speaking here's where it's at: score age size name 44.2 97 large GLM-5 (Reasoning) 44.7 187 - GPT-5.1 (high) 44.9 29 - Qwen3.6 Max Preview 45 0 - Gemini 3.5 Flash 45.5 27 large MiMo-V2.5-Pro 45.6 75 - GPT-5.4 (low) this is from artificial-analysis using https://github.com/day50-dev/aa-eval-email/blob/main/art-ana... I really don't know why people down vote me. What d…

I see no 'score' or 'age' mentioned in your script. What does age signify and how are they calculated?

This isn't obvious?

    "\(
        10 \* (.codingIndex // 0) | round / 10
    ) \(
      (
        now - (
        .releaseDate |
          try ( strptime("%Y-%m-%d") | mktime )
          catch (now + 86400)
      ) ) / 86400 | floor
Real question. I see 86400 and I know it's time... That might just be me.

I'm not being an ass, I don't know how to talk to people or when I think I'm being clear but I'm actually being cryptic

Re: Gemini 3.5 Flash

#423

Post by sho_hn »

airstrike wrote: post #386
hedgehog wrote: post #136

Earlier quoted context omitted.

That pelican looks like it's in Miami for a crypto conference.

They're called ClawCons now

Personally, I don't attend them since I figured out I can set up agents to performatively engage in AI-related discussion and events for me, freeing up tons of my time thanks to automation.

Truly: Nothing better than AI tools to brave the challenges and requirements of modern life. "Claude, ride the hype train" is the decisive prompt you need.

Re: Gemini 3.5 Flash

#424

Post by s3p »

brianwawok wrote: post #333
npn wrote: post #83

The price is crazy. And I guess Gemini 3.5 pro will have the pricing increment, too. 12 x 5 = 60? It seems like google does want us to use Chinese models.

What exactly are you doing with this that you can’t generate $1.50 of value per million tokens?

Wrong question.

Right question: What exactly is Google's plan for the long term pricing of these models, and are we all going to be priced out in a year?

Re: Gemini 3.5 Flash

#426

Post by p1esk »

alexandre_m wrote: post #414
lanthissa wrote: post #378

Earlier quoted context omitted.

no it really not, even the soggiest bank has multiple api vendors atm.

I agree with parent. I'm not sure where your stance is coming from. From what I hear, most enterprise AI deployments are seat-based subscriptions with annual commitments.

Yes, I work at a 50 person startup and even here switching from CC to codex or cursor would be non-trivial for multiple reasons - not just the annual commitment.

Re: Gemini 3.5 Flash

#427

Post by kristopolous »

kridsdale3 wrote: post #400
kristopolous wrote: post #262

I have a tool to track these I've built Relatively speaking here's where it's at: score age size name 44.2 97 large GLM-5 (Reasoning) 44.7 187 - GPT-5.1 (high) 44.9 29 - Qwen3.6 Max Preview 45 0 - Gemini 3.5 Flash 45.5 27 large MiMo-V2.5-Pro 45.6 75 - GPT-5.4 (low) this is from artificial-analysis using https://github.com/day50-dev/aa-eval-email/blob/main/art-ana... I really don't know why people down vote me. What d…

Buddy, this tone may be why. We genuinely don't understand what your post is about. What is this tool? What are these numbers representative? Why are things sorted in that order? You haven't communicated really anything at all. I am interested, I'd like to understand. Write a more complete post, please.

Are you familiar with https://artificialanalysis.ai/leaderboards/models

The json on the page has a coding index result it hides from the table.

That's what this exposes. It's a sorting from the leading evals company on the coding index for basically every model that matters presented in an easy to parse format that you can feed into model routing harnesses in real time so, for instance, your agents can dynamically upgrade themselves to better models as they come out or cost optimize based on eval results.

I do stuff like this, give it away for free and it's either ignored or makes people angry...

I really wish I didn't piss people off with my sincerity but somehow it always goes down that way

I really appreciate your time thank you so much

Re: Gemini 3.5 Flash

#428

Post by SwellJoe »

sheeshkebab wrote: post #417
akulbe wrote: post #351

Earlier quoted context omitted.

Any chance you'd be willing to talk further about your setup? I have 2 x 3090s in a local machine, and I'm still left with questions about how best to use stuff locally.

You can only run heavily quantized models on all 3/4/5 rtx gpus (with 32gb or less vram) - and you probably want moe versions like Qwen 35b for this to run at speed somewhat comparable to Claude. It’s still not there to be honest but getting there. Personally I mess around with llama.cpp on m5 max with 128gb - it’s a decent setup to try various medium sized things, and runs llms surprisingly well without quantization…

Two 3090s is 48GB, so it's possible to run the 6-bit quantization comfortably, which is fine. It doesn't start to get notably dumber until lower than that. It won't be as fast as a hosted model, but dual 3090s will be comfortably fast for interactive use with the MoE version and not terrible to use with the dense model. I run the dense model at 8 bits on my dual Radeon V620 desktop machine, which I think would be slower than two 3090s, or at least not notably faster.

Re: Gemini 3.5 Flash

#429

Post by s3p »

paol_taja wrote: post #399

That pelican looks like it just sold a SaaS company and bought a bike because its therapist said it needed balance.

The pelican is ready to discuss increased synergies of bringing AI to all teams at the firm!

Re: Gemini 3.5 Flash

#430

Post by taurath »

simonw wrote: post #131

The pelican is a lot : https://github.com/simonw/llm-gemini/issues/133#issuecomment... Not a great bicycle though, it forgot the bar between the pedals and the back wheel and weirdly tangled the other bars. Expensive too - that pelican cost 13 cents: https://www.llm-prices.com/#it=11&ot=14403&sel=gemini-3.5-fl...

I can’t help but think that what AI is best at is convincing management that things it creates are full featured which reads to their brains as mature
Post reply on HN