Should the title here be 4.6 to 4.7 instead of the other way around?
Writing Opus 4.6 to 4.7 does make more sense for people who read left to right.
I’m impressed with anyone who can read English right to left.
Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
21–30 of 620 posts
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#22Does anyone know what changed in the tokenizer? Does it output multiple tokens for things that were previously one token?
It must, if it now outputs more tokens than 4.6's tokenizer for the same input. I think the announcement and model cards provide a little more detail as to what exactly is different
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#23is it really unthinkable that another oss/local model will be released by deepseek, alibaba, or even meta that once again give these companies a run for their money
Now that Anthropic have started hiding the chain of thought tokens, it will be a lot harder for them
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#24is it really unthinkable that another oss/local model will be released by deepseek, alibaba, or even meta that once again give these companies a run for their money
I'm betting on a company like Taalas making a model that is perhaps less capable but 100x as fast, where you could have dozens of agents looking at your problem from all different angles simultaneously, and so still have better results and faster.
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#25Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#26Subsidies don't last forever.
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#27i think it is quite clear that staying with opus 4.6 is the way to go, on top of the inflation, 4.7 is quite... dumb. i think they have lobotomized this model while they were prioritizing cybersecurity and blocking people from performing potentially harmful security related tasks.
Hey, Boris from the Claude Code team here. People were getting extra cyber warnings when using old versions of Claude Code with Opus 4.7. To fix it, just run claude update to make sure you're on the latest.
Under the hood, what was happening is that older models needed reminders, while 4.7 no longer needs it. When we showed these reminders to 4.7 it tended to over-fixate on them. The fix was to stop adding cyber reminders.
More here: https://x.com/ClaudeDevs/status/2045238786339299431
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#28is it really unthinkable that another oss/local model will be released by deepseek, alibaba, or even meta that once again give these companies a run for their money
Nothing is unthinkable, I could think of Transformers.V2 that might look completely different, maybe iterations on Mamba turns out fruitful or countless of other scenarios.
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#29AFAICT this uses a token-counting API so that it counts how many tokens are in the prompt, in two ways, so it's measuring the tokenizer change in isolation. Smarter models also sometimes produce shorter outputs and therefore fewer output tokens. That doesn't mean Opus 4.7 necessarily nets out cheaper, it might still be more expensive, but this comparison isn't really very useful.
Why is it not useful? Input token pricing is the same for 4.7. The same prompt costs roughly 30% more now, for input.
Re: Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
#30is it really unthinkable that another oss/local model will be released by deepseek, alibaba, or even meta that once again give these companies a run for their money
Qwen released a new model the same day (3.6). The headline was kind of buried by Anthropic's release, though.