Kosta Milovanovic

454 posts

Kosta Milovanovic

Kosta Milovanovic

@ildunari

Biomedical Engineering Grad Student @ Brown University

Katılım Mayıs 2023
624 Takip Edilen26 Takipçiler
Brent Traut
Brent Traut@btraut·
Now that I've joined the Codex team, it's so freakin' cool being able to fix paper cut bugs that I was running into before the switch. What paper cuts are you running into? I'll see what I can do.
English
234
9
598
41.2K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@chetaslua Ok but that’s with no load on the servers. How high is it gonna be on release?
English
1
0
1
1.1K
Chetaslua
Chetaslua@chetaslua·
🚨 Gemini 3.5 Flash speed numbers just dropped 3-9x faster than current Flash peak speeds over 900 tokens/sec capability hasn't dropped, actually improved parameter count estimated larger than expected video: full shooter game one-shot. holds up at speed. Also currently testing 3.5 Pro on Arena. Should I start posting Pro outputs now or wait for public launch? vote below 🫡
English
20
14
354
25K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@elonmusk Give loyal subs with over a year of super grok sub free three month trial period. Helps you get feedback and stats and helps us test it out and make it our daily driver.
English
1
0
1
35
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@uzairansar @sudoingX Honestly kinda hope they come out and adopt the app as official. Would prob need to be open sourced though for that to happen.
English
1
0
1
114
Uzi
Uzi@uzairansar·
@sudoingX the community is strong - we got it! 🫡
English
1
0
6
290
Uzi
Uzi@uzairansar·
New feature added: Now you'll be able to pick a default profile in settings. Shipping in next build. requested by @ildunari
Uzi tweet media
English
22
2
39
5.7K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@uzairansar quick note. Can you add profile and model not just models default in settings? Since it affects the actual config, changing it in the chat or setting might change the wrong profiles defaults for TUI too since it’s not isolated. Try to map profile and model together.
English
1
0
1
22
Uzi
Uzi@uzairansar·
Alright, Hermes Mobile public Testflight is under review 👀 Should get you fiends on this as soon as it's approved!
Uzi tweet media
English
28
1
136
50.5K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@just_kaz @MagicPathAI They still don’t support clicking to open files in app right? Not sure how they’d do it, prob have to send each file that’s linked to some local db or cache.
English
1
0
1
37
Justin Kazwell
Justin Kazwell@just_kaz·
Codex in the ChatGPT mobile app, with @MagicPathAI CLI...I just prepped a sprint room while walking my dog.
Justin Kazwell tweet mediaJustin Kazwell tweet mediaJustin Kazwell tweet media
English
2
1
8
1.5K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@emanueledpt You’ll almost always loose against similar but free. My recommendation - turn the app into a hub for agents.
English
0
0
1
11
Emanuele Di Pietro
Emanuele Di Pietro@emanueledpt·
I promised to share my wins and losses Last 24 hours have been about loss. I've lost over 120 active subscribers, resulting in losing 1/10 of the total users. I'm bleeding right now, and I don't blame people for leaving The only thing that keeps me going are the ones that stayed saying that Remodex is still better I like to think that this situation will be temporary, because now, the only way to go is up. I'll keep losing subscribers in the next upcoming days, I know, probably I will end up with 50% of what I had at my peak. But I won't give up. That's why today I announced the creation of the official Remodex X account: @useRemodex This is a statement: Remodex will live. Remodex will be born again.
Emanuele Di Pietro tweet media
OpenAI@OpenAI

You've been asking for this one... Now in preview: Codex in the ChatGPT mobile app. Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox.

English
84
3
278
38.5K
Uzi
Uzi@uzairansar·
@erfan_bashar99 just sharing the journey as i build it man 😭
English
1
0
2
291
Uzi
Uzi@uzairansar·
new feature added to Hermes Mobile. share links/photos/files directly to Hermes Agent from the share sheet:
Uzi tweet media
English
12
1
108
8.8K
SIGKITTEN
SIGKITTEN@SIGKITTEN·
@doodlestein no why, i've been waiting so I can use those new APIs in my app lol. i've got a ton more ideas and experiments for kitty, it's early af
English
2
0
31
676
Stephen
Stephen@srotimi_ui·
Terminal component UI
Stephen tweet media
Català
22
21
614
18.8K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@dxrulezsrb @lydiahallie I’d say they’re in a slightly different boat. They don’t train nor own their models. And the OpenAI and Microsoft divorce is prob costing them inference. I’d put them in the same boat as cursor or other code agents serving expensive models.
English
1
0
1
35
Lydia Hallie ✨
Lydia Hallie ✨@lydiahallie·
To add some clarity: you don't pay extra. It's the same subscription, same price per month. What's new our sub now covers two separate pools: · Interactive → sub limits, unchanged · Programmatic → new $20–$200 included(!!) credit, metered at API rates
Lydia Hallie ✨ tweet media
ClaudeDevs@ClaudeDevs

Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK

English
346
53
771
541.7K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@dxrulezsrb @lydiahallie Except now that Anthropic did it first, the others are gonna follow suit in a few months. It’s always the first one that takes the hit, the rest of em follow soon after.
English
1
0
1
62
Milorad
Milorad@dxrulezsrb·
@lydiahallie This is why people switching to Codex and you are losing in your own game
English
1
0
3
1.4K
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@MrCollison @lydiahallie I mean they’re not giving it away. You’re paying for it. And you’re using the limits - whether in Claude first party or Third party. Why would it matter if I’m using Claude Cowork or Craft Agent via Agent SDK if they both hit the same usage? Why downgrade and force first party?
English
0
0
3
215
Matthew Collison
Matthew Collison@MrCollison·
I find it wild how many people are hit by this. I guess I never tried to use my key in another harness This new model is 100% reasonable - there's no biz model in giving $2k in API away for a $200 a month subscription In Claude Code it's different, because the data is valuable For me I just realised me and my dev team now have $200 mo extra usage to spend on testing other harnesses
English
18
1
27
7.4K
ClaudeDevs
ClaudeDevs@ClaudeDevs·
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
English
1.3K
1K
12.5K
10.2M
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@aronprins @ClaudeDevs Nah I’m complaining. It’s actually kind of shit. Claude apps are fine. But Craft Agent > Cowork for me. Idk lots of things I built around the SDK and now it gets only $200 credit a month.
English
1
0
3
248
Aron Prins
Aron Prins@aronprins·
@ildunari @ClaudeDevs Ahhhh…. So we get an “extra” $200 in credits to use in third party apps in a separate bucket? That’d ok-ish, but ruins claude -p for e.g. ralph loops which become extremely expensive fast this way. Not complaining, just want to make sure I fully understand :)
English
2
0
4
817
Kosta Milovanovic
Kosta Milovanovic@ildunari·
@aronprins @ClaudeDevs I’m pretty sure all native Claude apps go via usage. All non native apps (OpenClaw, T3, Agent SDK based stuff) uses the $200 monthly credit. When u run out, u still have ur native usage limits. The $200 is its own bucket, like how design is. And $200 is way less than x5.
English
1
0
3
1K
Aron Prins
Aron Prins@aronprins·
@ClaudeDevs Is this what you are saying: $200 gets you approx. 25x usage when we use your harnasses: Claude AI, Claude Code, Claude Desktop, etc. Or $200 gets us $200 in API credits for other harnasses. Do we need to choose between these two options?
English
8
0
26
16K