ntcho

10 karmajoined 5 jaar geleden

Submissions

OpenCode Data: Real-world AI model usage, cache ratios, and costs

2 pointsby ntcho3 dagen geleden1 comments

comments

3 dagen geledendiscuss

Came across this dashboard tracking live LLM usage on OpenCode users. Interesting production metrics, including token volume, session costs, and cache hit ratios. DeepSeek v4 currently dominates their market share.

ntcho

11 dagen geledendiscuss

Gosh, the last sentence is the most terrifying thing I’ve read today.

ntcho

vorige maanddiscuss

Average Firefox experience right here

ntcho

2 maanden geledendiscuss

absolutely love the domain here. great taste

ntcho

2 maanden geledendiscuss

I landed on a similar workflow. I found letting agents navigate / self-update docs on a repo full of markdown files quite successful

ntcho

2 maanden geledendiscuss

username checks out

ntcho

2 maanden geledendiscuss

Have you tried MacsFanControl? MBP fans rarely turn on by themselves, but controlling yourself makes throttling not a problem

ntcho

3 maanden geledendiscuss

I’m also having a blast with generating userscripts/css to make web SaaS work better for my workflows. Went from waiting for months for a new feature to writing QoL improvesments myself

ntcho

4 maanden geledendiscuss

I also noticed Gemini web UI failing to send messages on Firefox. Anyone else?

ntcho

6 maanden geledendiscuss

I think this should be on HN hall of fame: https://hn.algolia.com/?q=sabotage+manual

ntcho

8 maanden geledendiscuss

Decades of progress made in hardwares and networks… just to be negated with 100kb+ webpages

ntcho

9 maanden geledendiscuss

This is hilarious, thanks for sharing. Kinda crazy how well it works and already better than some ads

ntcho

9 maanden geledendiscuss

Has some quirks, but Beeper is exactly what you are describing: https://www.beeper.com/

ntcho

9 maanden geledendiscuss

This reminds me of cleaning a toaster with a dishwasher: https://news.ycombinator.com/item?id=41235662