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

GPT-5.5

openai.com

Post reply on HN
Print view View original

61–70 of 1001 posts

Re: GPT-5.5

#61

Post by conradkay »

jumploops wrote: post #28

> GPT‑5.5 improves on GPT‑5.4’s scores while using fewer tokens. This might be great if it translates to agentic engineering and not just benchmarks. It seems some of the gains from Opus 4.6 to 4.7 required more tokens, not less. Maybe more interesting is that they’ve used codex to improve model inference latency. iirc this is a new (expectedly larger) pretrain, so it’s presumably slower to serve.

They say latency is the same as 5.4 and 5.5 is served on GB200 NVL72, so I assume 5.4 was served on hopper.

Re: GPT-5.5

#62

Post by wahnfrieden »

ativzzz wrote: post #7

I like that they waited for opus 4.7 to come out first so they had a few days to find the benchmarks that gpt 5.5 is better at

I like that Anthropic rushed 4.7 out to get a couple days of coverage before 5.5 hit

Re: GPT-5.5

#63

Post by CSMastermind »

0x62 wrote: post #51
astlouis44 wrote: post #9

A playable 3D dungeon arena prototype built with Codex and GPT models. Codex handled the game architecture, TypeScript/Three.js implementation, combat systems, enemy encounters, HUD feedback, and GPT‑generated environment textures. Character models, character textures, and animations were created with third-party asset-generation tools The game that this prompt generated looks pretty decent visually. A big part of th…

FWIW I've been experimenting with Three.js and AI for the last ~3 years, and noticed a significant improvement in 5.4 - the biggest single generation leap for Three.js specifically. It was most evident in shaders (GLSL), but also apparent in structuring of Three.js scenes across multiple pages/components. It still struggles to create shaders from scratch, but is now pretty adequate at editing existing shaders. In 5.2…

> It still struggles to create shaders from scratch

Oh just like a real developer

Re: GPT-5.5

#64

Post by beering »

jumploops wrote: post #28

> GPT‑5.5 improves on GPT‑5.4’s scores while using fewer tokens. This might be great if it translates to agentic engineering and not just benchmarks. It seems some of the gains from Opus 4.6 to 4.7 required more tokens, not less. Maybe more interesting is that they’ve used codex to improve model inference latency. iirc this is a new (expectedly larger) pretrain, so it’s presumably slower to serve.

With Opus it’s hard to tell what was due to the tokenizer changes. Maybe using more tokens for the same prompt means the model effectively thinks more?

Re: GPT-5.5

#65

Post by Someone1234 »

I'd like to draw people's attention to this section of this page:

https://developers.openai.com/codex/pricing?codex-usage-limi...

Note the Local Messages between 5.3, 5.4, and 5.5. And, yes, I did read the linked article and know they're claiming that 5.5's new efficient should make it break-even with 5.4, but the point stands, tighter limits/higher prices.

Re: GPT-5.5

#66

Post by sosodev »

I hope the industry starts competing more on highest scores with lowest tokens like this. It's a win for everybody. It means the model is more intelligent, is more efficient to inference, and costs less for the end user.

So much bench-maxxing is just giving the model a ton of tokens so it can inefficiently explore the solution space.

Re: GPT-5.5

#67

Post by ionwake »

is there anywhere I can try it? ( I just stopped my pro sub ) but was wondering if there is a playground or 3rd party so i can just test it briefly?

Re: GPT-5.5

#70

Post by vexna »

k2xl wrote: post #44

Surprised to see SWE-Bench Pro only a slight improvement (57.7% -> 58.6%) while Opus 4.7 hit 64.3%. I wonder what Anthropic is doing to achieve higher scores on this - and also what makes this test particular hard to do well in compared to Terminal Bench (which 5.5 seemed to have a big jump in)

There's an asterisk right below that table stating that:

> *Anthropic reported signs of memorization on a subset of problems

And from the Anthropic's Opus 4.7 release page, it also states:

> SWE-bench Verified, Pro, and Multilingual: Our memorization screens flag a subset of problems in these SWE-bench evals. Excluding any problems that show signs of memorization, Opus 4.7’s margin of improvement over Opus 4.6 holds.

Post reply on HN