ex yc software eng. currently building: Ness, #1 personal AI Health Coach Atom, coding agent in Rust

Based in United States
Filter
Exclude
Time range
-
Minimum likes
Replying to @Xxi5olc @scottstts
API spend varies greatly from subscription spend, for example a lot more people use Fable over subscription than api
1
14
Replying to @lukasjohnson
Hey Lukas! DM me and I'll send over a code for a free year of Ness
1
8
Replying to @RajitWrites
Congrats on the launch! Are there plans to add 5.3-flash shortly in the future?
1
1
92
Replying to @skalskip92
does astra really provide a big enough perf boost for the spend? how does opus 5.5 and 3.8 flash compare?
1
138
Replying to @Ananth7e
Drake
4
Replying to @iamdavidhill
I really hope the numbers are significantly better on release, otherwise it doesn't make much sense to use go plus
82
Replying to @pvncher
I'm happy that the 6-series improved on cost, but saying it's a flatly better model isn't completely accurate Here's an example of a benchmark where it underperforms compared to the prior versions
69
yes over-the-air updates exist, but you still need to submit for review anytime you add native code
9
Apple needs to hire more app reviewers fast. On web you can deploy in minutes, on iOS it takes multiple days, depending on when you get your review. Increasingly, one bottleneck to shipping quality updates becomes review time..
2
30
it seems like Instinct finally had to add usage limits, makes sense you NEED limits if you're serving high quality models, and aren't a huge company like Meta/OAI
17
Replying to @thesamparr
would love for you to try Ness, it's an all-in-one AI Health coach with nutrition built in. You can text it photos too if you'd like happy to send over a lifetime code so you can try it out!
35
Replying to @chetaslua
that's kind of sad if true, it seems to perform worse than minimax m3
109
Replying to @sheriyuo
Wdym, ox alpha was glm 5.3 flash Minimax is way overdue on model releases so it would make sense it’s them
4
602
Do you have plans to list on Vercel's AI Gateway? We need more fast, open-weight models on there
2
1
82
Claude Opus 5.5 has the best visual design of any model I have tested so far
Claude-Pop - I'm Upping My P(Doom)
270
690
6,594
2,398,775
Replying to @KeyTryer
Why is it kind of good
1
834
Replying to @Da7_Tech
depends on the inference providers you use and the laws/regulations they follow
1
265
True intelligence too cheap to meter
2
20
Replying to @davis7
ok what, 3.8 flash is at least ASI
209
Replying to @chribjel
Kimi had it before grok bot, besides it's just a common mascot 😆
103