CryptoPraetoria

@jpschroeder Nice I am very curious

@CryptoPraetoria Ill do an OSS model roundup at some point too.

Hold on. What? No. What? Is Composer 2.5 the best model at UI now? "Create a dog walking service landing page." GPT 5.5 in Codex vs Opus 4.7 in Claude Code vs Composer 2.5...

@SunFoxx_ @thsottiaux Yeah that’s absolute true

@thsottiaux The limits overconsumption is still not fixed though, and the model has become noticeably dumber than a week ago. Seems like a general consensus across many users. Any plans to address this?

There is a certain zen to looking at codex traffic, usage and compute dashboards late at night while listening to LCD Soundsystem. The tokens must flow

@thsottiaux Please do a manual usage reset

@minchoi Design?

This is literally my new workflow now: Real-time research/search → Grok 4.3 Planning & Reasoning → Codex w/ GPT-5.5 Coding → Codex w/ GPT-5.5 Write Test Cases → Codex w/ GPT-5.5 Run Test Cases → Codex w/ GPT-5.5 Debug → Codex w/ GPT-5.5 Need Mythos & Grok Build on V9...

@jturntdev Yes it was way more intelligent at the beginning

Is it just me or ChatGPT 5.5 nerfed? This isn’t the model i used on release. Claude’s somehow seeing cracks that 5.5 is completely missing.

@Aaronbuilds_ @rezoundous How they talk to each other?

@rezoundous Nah... A $100 Claude talking with a $100 Codex is way better than a single $200 Codex or Claude.

It's honestly so freeing to not have multiple AI subscriptions.

@thsottiaux @rezoundous When good ui design with codex?

@rezoundous Congrats, you figured out 5% of our roadmap

OpenAI model wishlist. 1. Good limits 2. Good UI design That's all.

@rileybrown @Allan_Ryan_ Why not?

@Allan_Ryan_ Nah

@rileybrown Hermes

I actually can’t stand using AI tools that don’t have extensive context of me and my goals. I tried to use Notions built in agent features and immediately got mad when it didn’t have context over me. I’m using 2 AI tools now > Codex and > a Claw through iMessage. Claw = Persistent agent running on its own computer 24/7, with memory that I can text (Not OpenClaw its been unreliable)

@thsottiaux When computer use for Europe?

Using computer use, you can ask codex to cancel subscriptions you don't need anymore. Very pleasant to watch. No particular one in mind, works on all of them. chatgpt.com/codex/

@shannholmberg Is it better than obsidian?

What´s gBrain and how does it work? I've been using gStack for a while when ideating, validating new projects, and some coding now I'm experimenting with gBrain as the memory layer for my agents, starting with my Hermes Agent company gBrain is an open-source persistent memory layer for AI agents (by @garrytan). it turns your emails, meetings, tweets, voice memos, and docs into a typed knowledge graph. essentially markdown in, graph out. how it works: > 1. ingest signals from your daily life > 2. extract entities + create typed links (works_at, invested_in, attended) > 3. store as Markdown + Postgres + pgvector > 4. retrieve via hybrid search (keyword + vector + graph) > 5. agents read brain first, write insights back, graph builds itself an overnight dream cycle dedupes entities, repairs links, and updates the compiled truth

@maxjendrall Have to try this

Codex too that /goal very serious haha

@pashmerepat Sounds very interesting tho

have a long running personal finance thread in the codex app set up with a heartbeat automation. as part of it, I gave Codex access to all my banks, credit cards, tax statements & brokerages. also built an agent-first Schwab CLI so it can view my holdings and open up trades. just asked codex to open source it. woke up this morning and saw this notification on my phone — codex made some trades (: at this point, codex knows my financial picture better than I do, and I use this pinned thread as my primary interface for anything money related. obviously handing over all your finances to ai has risks, but personally i’m stoked about this. feels good having codex in my corner helping me stay on top of frivolous spending, taxes, and portfolio

@0xRemakw @thsottiaux Limits getting crushed since couple days

CODEX LIMITS REDUCED 50% can someone verify if this is true ? @thsottiaux

@JSFILMZ0412 Because omni is for editing videos and not for creating

Google promised that Gemini Omni Flash would change the game, but when you put it side-by-side with Seedance 2.0... it’s not even a fair fight. Fingers crossed for Veo4

@SqueakAlGaib @heygentlewhale @victor_vibing Hahahaha 😂😂😂

@heygentlewhale @victor_vibing They had 1 year to cook. And they cooked trash.

Kling3 vs Seedance2 vs Omni Flash side by side test. hmmm...

@adam_fresko @morganlinton No you can use it outside

@morganlinton Composer is only in the Cursor app, right? I need API access.

Now I'm getting confused. Does Gemini 3.5 Flash suck at coding, or is it amazing? 😵‍💫

@jpschroeder @leerob Yeah I want it too

@leerob give us an API and put it on openrouter.

Where could we improve Composer 2.5? We're working on the next model and would love your feedback. Lots of work to do (our CursorBench evals below) in the coming weeks!

@DannyLimanseta @threejs Do you use it as ide inside cursor?

Very impressed with Composer 2.5 after about a day of usage. I've almost moved over to it exclusively from GPT 5.5, even using it for planning now. It's like Opus 4.7 on steroids, crazy fast. Fast models really get me into the flow of building, which is exhilarating. Also, a sneak peek of a weekend mini-project I'm building: a racing game where you race your mouse cursor.