frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

I cut Claude API costs from $70/month to pennies

2•ok_orco•1h ago
The first time I pulled usage costs after running Chatter.Plus - a tool I'm building that aggregates community feedback from Discord/GitHub/forums - for a day hours, I saw $2.30. Did the math. $70/month. $840/year. For one instance. Felt sick.

I'd done napkin math beforehand, so I knew it was probably a bug, but still. Turns out it was only partially a bug. The rest was me needing to rethink how I built this thing. Spent the next couple days ripping it apart. Making tweaks, testing with live data, checking results, trying again. What I found was I was sending API requests too often and not optimizing what I was sending and receiving.

Here's what moved the needle, roughly big to small (besides that bug that was costin me a buck a day alone):

- Dropped Claude Sonnet entirely - tested both models on the same data, Haiku actually performed better at a third of the cost

- Started batching everything - hourly calls were a money fire

- Filter before the AI - "lol" and "thanks" are a lot of online chatter. I was paying AI to tell me that's not feedback. That said, I still process agreements like "+1" and "me too."

- Shorter outputs - "H/M/L" instead of "high/medium/low", 40-char title recommendation

- Strip code snippets before processing - just reiterating the issue and bloating the call

End of the week: pennies a day. Same quality.

I'm not building a VC-backed app that can run at a loss for years. I'm unemployed, trying to build something that might also pay rent. The math has to work from day one.

The upside: these savings let me 3x my pricing tier limits and add intermittent quality checks. Headroom I wouldn't have had otherwise.

Happy to answer questions.

Comments

arthurcolle•1h ago
Can you discuss a bit more of the architecture?
ok_orco•1h ago
Pretty straightforward. Sources dump into a queue throughout the day, regex filters the obvious junk ("lol", "thanks", bot messages never hit the LLM), then everything gets batched overnight through Anthropic's Batch API for classification. Feedback gets clustered against existing pain points or creates new ones.

Most of the cost savings came from not sending stuff to the LLM that didn't need to go there, plus the batch API is half the price of real-time calls.

Ask HN: Some great launch videos in recent times?

1•nemath•2m ago•0 comments

Should I unsubscribe from Shane Parrish's AI-generated newsletter?

1•arnz-arnz•3m ago•0 comments

Microsoft suspects some PCs might not boot after Windows 11 January 2026 Update

https://www.windowslatest.com/2026/01/25/microsoft-suspects-some-pcs-might-not-boot-after-windows...
2•nsoonhui•4m ago•0 comments

Show HN: Nhx – Node.js Hybrid eXecutor (a uvx inspired tool)

https://www.npmjs.com/package/nhx
1•kolodny•5m ago•0 comments

Show HN: Endfield Calculator – Arknights Factory and Base Planning Tool

https://endfieldcalculator.org/
2•tomstig•7m ago•0 comments

KASM Workspaces

https://docs.linuxserver.io/images/docker-kasm/
1•indigodaddy•10m ago•0 comments

Video Games as Art

https://gwern.net/video-game-art
2•andsoitis•13m ago•0 comments

Show HN: Debugging conflicting U.S. sexual behavior surveys

https://osf.io/preprints/socarxiv/jcdbm_v2
2•joshuafkon•13m ago•0 comments

Show HN: Invoice Studios – local first invoicing app (one time purchase)

https://liblab.gumroad.com/l/invoice-studio
1•josephttd•15m ago•0 comments

Show HN: Interactive demo of the X "For You" algorithm (runs in browser)

https://prabal.ca/x-algorithm/
3•prabal97•17m ago•0 comments

SF Microclimates API

https://github.com/solo-founders/sf-microclimates
1•weisser•19m ago•0 comments

AI FOMO

https://datamethods.substack.com/p/ai-fomo
1•zekrom•20m ago•1 comments

Show HN: Cloister – Local web UI to browse and monitor Claude Code sessions

https://github.com/bradleyboy/cloister
1•bradleyboy•21m ago•0 comments

Pragtical Editor v3.8.2 Released

https://pragtical.dev/blog/pragtical-v382-release
2•rd07•28m ago•0 comments

Show HN: Omiu.me – A WYSIWYG profile builder that uses blocks

https://omiu.me/
1•zacaryn•30m ago•0 comments

Why is cursor / Claude Code is so bad at generating readmes?

2•yakshithk_•37m ago•1 comments

Vectorized MAXSCORE over WAND, especially for long LLM-generated queries

https://turbopuffer.com/blog/fts-v2-maxscore
1•vismit2000•41m ago•0 comments

How Google AI Overviews are putting public health at risk

https://www.theguardian.com/technology/ng-interactive/2026/jan/24/how-the-confident-authority-of-...
2•oenton•44m ago•0 comments

No Politics on Hacker News

https://joelx.com/no-politics-on-hacker-news/
5•silexia•45m ago•5 comments

Learning with LLMs

https://jwuphysics.github.io/blog/2025/12/learning-with-llms/
2•jxmorris12•46m ago•0 comments

Staggering Beauty 2

https://staggeringbeauty.io/
1•jackisguess•46m ago•0 comments

LLMs Aren't Tools

https://yagmin.com/blog/llms-arent-tools/
1•lubujackson•47m ago•0 comments

Sysp: Systems Lisp compiling to C with homoiconic macros, refcounted memory, Hi

https://github.com/karans4/sysp
1•todsacerdoti•48m ago•0 comments

The largest Trump superPAC donor so far this cycle is the president of OpenAI

https://bsky.app/profile/jakemgrumbach.bsky.social/post/3mdbzv2nfsc2k
16•m-hodges•49m ago•0 comments

Show HN: A Local OS for LLMs. MIT License. Zero Hallucinations. Infinite Memory

https://github.com/merchantmoh-debug/Remember-Me-AI
1•MohskiBroskiAI•51m ago•0 comments

Recursive Language Models: the paradigm of 2026

https://www.primeintellect.ai/blog/rlm
1•pseudolus•52m ago•0 comments

How Revolutions Really Start

https://neilthanedar.com/how-revolutions-really-start/
2•thanedar•55m ago•1 comments

Monster Neutrino Could Be a Messenger of Ancient Black Holes

https://www.quantamagazine.org/monster-neutrino-could-be-a-messenger-of-ancient-black-holes-20260...
1•pseudolus•58m ago•0 comments

Show HN: BSS Blue Hive Build

https://www.bluehiveguide.com/blue-hive-composition-guide.html
1•andy846851797•58m ago•0 comments

zerobrew is a 5-20x faster, Rust-based homebrew replacement

https://github.com/lucasgelfond/zerobrew
3•lucasgelfond•59m ago•0 comments