News for hackers. Stuff that matters.

A comment on We cut Codex's input token cost by 49.5% with a compression gateway (benchmark).

Re:We cut Codex's input token cost by 49.5% with a compressi... by MallocVoidstar on Thursday April 09, 2026 @01:20PM (#47706440)
No discussion on problem difficulty, or on result quality besides "the Edgee run generated slightly more output tokens than the baseline".
[ Parent | on HN ]
Re:We cut Codex's input token cost by 49.5% with a compressi... by sachamorard on Thursday April 09, 2026 @01:32PM (#47706645)
More info in the GitHub repo, in the reports folder (sorry, I'm not sure I can add the link here without being flagged).

"Codex + Edgee consumes roughly half the fresh tokens of the normal Codex baseline. Output tokens are marginally higher (+3,312, +19.5%), suggesting the Edgee scenario produces slightly more verbose responses but dramatically reduces context ingestion."

[ Parent | on HN ]
Re:We cut Codex's input token cost by 49.5% with a compressi... by kokakiwi on Thursday April 09, 2026 @01:38PM (#47706736)
I think the problem being given to Codex for the benchmark is the one in the attached video, where two Codex run side-by-side, working a "standard" dev thingy
[ Parent | on HN ]