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

Measuring Claude 4.7's tokenizer costs

claudecodecamp.com

Post reply on HN
Print view View original

81–90 of 540 posts

Re: Measuring Claude 4.7's tokenizer costs

#81

Post by arcanemachiner »

WarmWash wrote: post #72
devmor wrote: post #58

Earlier quoted context omitted.

Skeptical is a light way to put it. It is essentially a forgone conclusion that once a company IPOs, any veil that they might be working for the global good is entirely lifted. A publicly traded company is legally obligated to go against the global good.

The problem is that people equate money to power and power to evil. So no matter what, if you do something lots of people like (and hence compensate you for), you will be evil. It's a very interesting quirk of human intuition.

A reasonable conclusion, considering that money and power seem to have their own gravity, so people with more of both end up getting even more of both, and vice versa.

Can't blame someone who comes to such a conclusion about money and power.

Re: Measuring Claude 4.7's tokenizer costs

#82

Post by robot_jesus »

ZeroCool2u wrote: post #71
_pdp_ wrote: post #29

IMHO there is a point where incremental model quality will hit diminishing returns. It is like comparing an 8K display to a 16K display because at normal viewing distance, the difference is imperceptible, but 16K comes at significant premium. The same applies to intelligence. Sure, some users might register a meaningful bump, but if 99% can't tell the difference in their day-to-day work, does it matter? A 20-30% cost…

Whenever we get the locally runnable 4k models things are going to get really awkward for the big 3 labs. Well at least Google will still have their ad revenue I guess.

They're not perfect but the local model game is progressing so quickly that they're impossible to ignore. I've only played around with the new qwen 3.6 models for a few minutes (it's damn impressive) but this weekend's project is to really put it through its paces.

If I can get the performance I'm seeing out of free models on a 6-year-old Macbook Pro M1, it's a sign of things to come.

Frontier models will have their place for 1) extensive integrations and tooling and 2) massive context windows. But I could see a very real local-first near future where a good portion of compute and inference is run locally and only goes to a frontier model as needed.

Re: Measuring Claude 4.7's tokenizer costs

#83

Post by mattkevan »

devmor wrote: post #58
giwook wrote: post #52

Earlier quoted context omitted.

I agree, and I'm also quite skeptical that Anthropic will be able to remain true to its initial, noble mission statement of acting for the global good once they IPO. At that point you are beholden to your shareholders and no longer can eschew profit in favor of ethics. Unfortunately, I think this is the beginning of the end of Anthropic and Modei being a company and CEO you could actually get behind and believe that…

Skeptical is a light way to put it. It is essentially a forgone conclusion that once a company IPOs, any veil that they might be working for the global good is entirely lifted. A publicly traded company is legally obligated to go against the global good.

It’s not really, companies like GM used to boast about how well they treated their employees and communities. It was Jack Welch and a legion of like-minded arseholes who decided they should be increasingly richer no matter who or what paid for it.

Re: Measuring Claude 4.7's tokenizer costs

#84

Post by therobots927 »

adaptive_loop wrote: post #66

Every time a new model comes out, I'm left guessing what it means for my token budget in order to sustain the quality of output I'm getting. And it varies unpredictably each time. Beyond token efficiency, we need benchmarks to measure model output quality per token consumed for a diverse set of multi-turn conversation scenarios. Measuring single exchanges is not just synthetic, it's unrealistic. Without good cost/qua…

That’s the joy of purchasing an intangible and non-deterministic product. The profit margin is completely within the vendor’s control and quality is hard for users to measure.

Re: Measuring Claude 4.7's tokenizer costs

#86

Post by aray07 »

namnnumbr wrote: post #56

The title is a misdirection. The token counts may be higher, but the cost-per-task may not be for a given intelligence level. Need to wait to see Artificial Analysis' Intelligence Index run for this, or some other independent per-task cost analysis. The final calculation assumes that Opus 4.7 uses the exact same trajectory + reasoning output as Opus 4.6. I have not verified, but I assume it not to be the case, given…

im running some experiments on this but based on what i have seen on my own personal data - I dont think this is true

"given that Opus 4.7 on Low thinking is strictly better than Opus 4.6 on Medium, etc., etc.”

Opus 4.7 in general is more expensive for similar usage. Now we can argue that is provides better performance all else being equal but I haven’t been able to see that

Re: Measuring Claude 4.7's tokenizer costs

#87

Post by unpwn »

namnnumbr wrote: post #56

The title is a misdirection. The token counts may be higher, but the cost-per-task may not be for a given intelligence level. Need to wait to see Artificial Analysis' Intelligence Index run for this, or some other independent per-task cost analysis. The final calculation assumes that Opus 4.7 uses the exact same trajectory + reasoning output as Opus 4.6. I have not verified, but I assume it not to be the case, given…

Very unlikely that the article is wrong. the 4.7 intelligence bump is not that big, plus most of the token spend is in inputs/tool calls etc, much of which won't change even with this bump.

Re: Measuring Claude 4.7's tokenizer costs

#89

Post by bakugo »

beej71 wrote: post #65

News like this always makes me wonder about running my own model, something I've never done. A couple thousand bucks can get you some decent hardware, it looks like, but is it good for coding? What is your all's experience? And if it's not good enough for coding, what kind of money, if any, would make it good enough?

You should be aware that any model you can run on less than $10k worth of hardware isn't going to be anywhere close to the best cloud models on any remotely complex task.

Many providers out there host open weights models for cheap, try them out and see what you think before actually investing in hardware to run your own.

Re: Measuring Claude 4.7's tokenizer costs

#90

Post by sysmax »

gruez wrote: post #74
sysmax wrote: post #73

Well, LLMs are priced per token, and most of the tokens are just echoing back the old code with minimal changes. So, a lot of the cost is actually paying for the LLM to echo back the same code. Except, it's not that trivial to solve. I tried experimenting with asking the model to first give a list of symbols it will modify, and then just write the modified symbols. The results were OK, but less refined than when it e…

>and most of the tokens are just echoing back the old code with minimal changes I thought coding harnesses provided tools to apply diffs so the LLM didn't have to echo back the entire file?

They can, but this reduces the quality. The LLM has a harder time picking the first edit, and then all subsequent work is influenced by that one edit. Like first creating an unnecessary auxiliary type, and then being stuck modifying the rest of the code to work with it.

So, in practice, many tools still work on the file level.

Post reply on HN