kokakiwi
Compressor V2: three compression layers for a 50% LLM agent cost cut
Re:Your hex editor should color-code bytes
on Thursday April 23, 2026 @06:58AM (#47874234)
ImHex (https://imhex.werwolv.net/) is also a really nice Hex editor with tons of plugins (patterns, file support, etc.) and even an embedded language for adding more patterns easily
Re:Claude Opus 4.7
on Thursday April 16, 2026 @12:11PM (#47795523)
Headroom looks great for client-side trimming. If you want to tackle this at the infrastructure level, we built Edgee (https://www.edgee.ai) as an AI Gateway that handles context compression, caching, and token budgeting across requests, so you're not relying on each client to do the right thing.
(I work at Edgee, so biased, but happy to answer questions.)
Re:We cut Codex's input token cost by 49.5% with a compression gateway...
on Thursday April 09, 2026 @01:47PM (#47706890)
For coding agents, mainly the tools' output, they're often the heaviest "messages" sent by the user and also the most "noisy" (like for "cargo test", Codex don't really care about all the build part, only the test results)
Re:We cut Codex's input token cost by 49.5% with a compression gateway...
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