> Claude Opus 4.7 and later Opus models, Claude Fable 5, Claude Mythos 5, Claude Mythos Preview, and Claude Sonnet 5 use a newer tokenizer that contributes to their improved performance on a wide range of tasks. This tokenizer produces approximately 30% more tokens for the same text. Claude Sonnet 4.6 and earlier models use the previous tokenizer.
Did anyone try to send a long email that pushed context close to the limit to try and make the agent a bit fuzzy on its original directive not to leak the secrets?
I don't know what the solution to this is, but I find it somewhat unfair that I pay money to Anthropic, and I pay money to OpenAI, and neither of them will let me use their best models for securing the software I work on.
Admittedly Opus 4.8 xhigh does a good job, but are my customers not entitled to have more security from a Fable/Mythos or GPT-5.5-Cyber audit over the codebase? Or I guess the inverse question: why aren't they allowed that audit?
(Fable/Mythos being unavailable notwithstanding.)
It seems OpenAI will at least let me do this narrowly, at greater cost, by using one of their partners. But I already pay them money!
OpenAI really snatched defeat from the jaws of victory late last year when Claude Code was a laggy mess.
Nowadays Codex has typing latency out of the gate, whereas Claude Code has the odd pause but generally displays my key presses as … you know … I press them.
Personally I don’t understand why Claude Code doesn’t have a mode to make text green and characters come down from the top of the screen individually, like in The Matrix.
I had a very bad start to it too, it lost track of where my source code was (in the repo! the current working directory!) and started grepping for .gitignore trying to get a foothold on where the git repo was.
And after that asked some questions that it already had answers to.
Started a brand new session and it's been OK since. Only drawn one silly conclusion so far, which I nudged it away from.
I came back to a workplace, that still used JIRA. Obviously during the interview I was like oh JIRA yeah yeah yeah you still use that? I can use that.
Anyway yes, I can use JIRA. But it was a real shock to see the latest version of JIRA. It has a thousand papercuts, one of the worst is double clicking on text select stuff suddenly kicks fields into editor mode.
What I was remembering was JIRA Server 4.0, you can walk down memory lane here* - zoom in enough and you'll see each issue has a title, type, fix version, affects version, and so on, and then you end up going straight to the comments. Very straightforward.
Anthropic seems to have a modest lead on their harness and models, so it’s a best-of-both-worlds scenario.
> I'm not sure what Microsoft is doing behind the scenes
It’s probably the exact same model, but the tools and the prompts around it are worse, so you get worse results.