Hacker News new | ask | show | jobs
by timcobb 3 days ago
Codex uses 256k masterfully, 1M is luxurious but still quite expensive and seems not necessary as a default.
4 comments

Needing large context windows is an illusion.
While I tend to clear my session after every task, I feel less stressed when my context is as 10% than when it's at 30%
Needing large context window is a skill issue
So, all types of knowledge work and coding require the same amount of context?

This seems like a hot take not really informed by experience.

This is a discussion on context window and not context.

Please read my essay on context window saturation. I am also the author of deepcq (intent-aware memory querying & context mgmt for LLM agents)

Please engage with my objection instead of assigning homework.
> Codex uses 256k masterfully

This.

When working on hard problems (not "vibecode me a script to show an alert box", but e.g. "let's see what this three-level LUT-state-machine obfuscated binary does"), hitting the 1M (!) context window with Claude Code feels like you were talking to Claude Claudewski when his shift just abruptly ends, he packs his things, throws the office keys at Claude Claudeson in-between the front door frame while handovering like "Hi! Nice to see you, good luck." and now here we go again, you are working with someone who just experienced an acute amnesia. It tries everything it already tried, everything it was told in the initial prompt to not do, everything it was told in follow-up prompts not to do. "You were right, this approach does not work and we don't have 20 TB RAM on this machine for full symbolic execution, let me try..."

In Codex, it's so seamless that I sometimes just notice "wait, the context was 20 % remaining, it is 70 % now, wow, when did this happen", while it seamlessly works on the task. Basically never had an issue with context on Codex, be it coding features, cracking hard crack-me ciphers, or researching basically anything.

(For full disclosure, my last experience with Claude was a few weeks ago when I cancelled the subscription, maybe they fully reworked the traumatic "Summarizing" - "Oh, hi! Where are we? Who am I? What we are doing? This is taking too long, let me take a shortcut..." lobotomy they were doing in the meantime.)

256k is enough when the harness uses it properly and the model is not stupid. And also when the tokenizer is not tuned to invoice as many tokens as possible...

I'll have to give it a shot some time soon. I just can't imagine working on any of my serious projects doing a new large feature with that little room. By the time it has looked at half the code required to start planning the work, it'd be out of context. Must use sub agents better or something.
Codex turns over 256k windows for *days/weeks on end* without seemingly much fidelity loss, especially with a goal that anchors it to a durable plan/ledger.
Also, a token means something different in each model/tokenizer. "For improved performance we will use (read: invoice you) 30 % more"...

https://platform.claude.com/docs/en/about-claude/models/what...

That's because it has the best compaction of anyone, and it's not even close.

Claude might as well not even do it in my experience.