Sign in

Melvynx

@melvynx.dev
167 followers 205 following 3.2K posts

I create content about React / Next.js / TypeScript online and Indie Hacker

PostsRepliesMedia
Melvynx @melvynx.dev · 2h
Claude: Can I do this? Me:
A text message discussing GitHub tasks, emphasizing commands like push, clean, and renaming repositories.
000
Melvynx @melvynx.dev · 11/10/2026
i hate hetzner now
000
Melvynx @melvynx.dev · 09/10/2026
New episode of our podcast with @ledevultime and I promise you that it will be insane. We talked about: * the myth of the replacement (it doesn't exist) * the end of tutorial videos * how to make a personal brand in the AI era You should watch it now: mlv.sh/p3?c=1
youtu.be
L'IA a TUÉ les tutoriels YouTube (et les créateurs… ?) ft. le dev ultime | E03
podcast 3
000
Melvynx @melvynx.dev · 06/10/2026
Opus 5.5 VS GPT 6 Astra for video editing (same skill, same ressources) Opus 5.5 cook so hard
000
Melvynx @melvynx.dev · 02/10/2026
For French speakers 🇫🇷, I just published my HARD CORE podcast with @YannisHaismann and @JohnLrDev where we talk about: • why developers have orgasms with Claude and Codex • why video editing is dead • open 500 subs, soon 1000 and then 10k subs? Link here: mlv.sh/lepodcastia...
youtu.be
IL A UN ORGAS** AVEC CLAUDE ? On a viré tous nos monteurs vidéo, OpenAI te fait payer 500 €
Tu ne rêves pas
010
Melvynx @melvynx.dev · 01/10/2026
I have too many Stripe account i guess
000
Melvynx @melvynx.dev · 30/09/2026
Sol is still terrible in motion design btw
000
Melvynx @melvynx.dev · 30/09/2026
unlimited Claude for $200 or limited Codex for $500 the dev day has been really terrible
010
Melvynx @melvynx.dev · 30/09/2026
Codex can now open a pull request from ChatGPT Work or the desktop app. Inspect the diff, ask about changes, and get feedback on your code before sharing it. GitHub review is slowly moving inside the agent. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
Open a pull request in the ChatGPT Work or Codex desktop app, inspect the diff, and ask Codex about the changes. You can also get feedback on your own changes before sharing them: https://t.co/M…
000
Melvynx @melvynx.dev · 30/09/2026
ChatGPT plugins are becoming actual mini-apps. You can build interactive panels, file viewers and sidebar homes. MCP Events also let connected apps trigger automations. This is much closer to an OS than a chatbot. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
Plugin extensions let you build interactive panels, file viewers, and a sidebar home in ChatGPT. Get clearer feedback when submitting your plugin for review. With support for the proposed MCP Ev…
100
Melvynx @melvynx.dev · 30/09/2026
OpenAI now has a $500/month plan. Pro 500 gives you 25x the Plus usage limits plus Ultrafast in Codex and ChatGPT Work. They also reopened Pro 200. AI subscriptions are becoming the new cloud bill. x.com/OpenAI/stat...
x.com
OpenAI (@OpenAI) on X
Ultrafast is available today for GPT-6 Astra in Codex, ChatGPT Work, and the API, with GPT-6.1 Sol coming soon. To access it in Codex and ChatGPT Work, we’re introducing Pro 500—a new plan with o…
100
Melvynx @melvynx.dev · 30/09/2026
You can finally close your laptop without killing your agents. Codex cloud environments keep tasks running with your repo, dependencies and settings preloaded. Less setup. Faster starts. Agents that actually finish. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
You can finally close your laptop now and your agents will keep working. Codex cloud environments are here. Reusable environments mean less setup and faster starts, with your repo, dependencies,…
100
Melvynx @melvynx.dev · 30/09/2026
The new Decisions API is built for fast routing and classification. Define the possible answers, then GPT-6 Luna picks one in ~150 ms. OpenAI’s demo routed 10,000 support requests around 10x faster. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
Give your app real-time decision-making with Decisions API, powered by GPT-6 Luna. Define questions and possible answers to classify content, route requests, or choose an agent’s next action. Av…
100
Melvynx @melvynx.dev · 30/09/2026
Astra Ultrafast is ridiculous. Up to 8x faster than Standard, 4x faster than Fast, and around 300 tokens/sec in Codex. It’s available in Codex, ChatGPT Work and the API for real-time AI experiences. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
Ultrafast is our fastest way to build with Astra yet: in Codex, it runs up to 8x faster than Astra Standard and 4x faster than Astra Fast. Bring your ideas to life as fast as you can type them.
100
Melvynx @melvynx.dev · 30/09/2026
OpenAI introduced Dots: always-on agents powered by GPT-6 Astra. Think OpenClaw with a polished consumer UX: 24/7 work, app connections, cloud computers and proactive tasks inside boundaries you control. x.com/OpenAI/stat...
x.com
OpenAI (@OpenAI) on X
Introducing dots, powered by GPT-6 Astra. Remarkably capable, always-on agents built to handle everything.
100
Melvynx @melvynx.dev · 30/09/2026
Codex CLI finally got the update it needed. Full-screen UI, better readability, parallel agents, /agents, /fork, voice, themes, Mermaid and LaTeX. Less “CLI wrapper”, more complete agent workspace. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
Codex CLI just got a major refresh, with a new look and powerful new capabilities. We’re continuing to invest in Codex CLI for developers who live in the command line. New full-screen interface,…
110
Melvynx @melvynx.dev · 30/09/2026
GPT-6.1 Sol is OpenAI’s answer to Opus 5.5. Better agentic coding + computer use, near-Astra performance, and cached inputs 95% cheaper. Built for huge refactors and agents that work for hours. x.com/OpenAIDevs/...
x.com
OpenAI Developers (@OpenAIDevs) on X
GPT-6.1 Sol is here. Upgraded with stronger agentic coding and computer use, near-Astra performance, and cached input at a 95% discount to standard input pricing. GPT-6.1 Sol is built for comple…
100
Melvynx @melvynx.dev · 30/09/2026
OpenAI just dropped enough updates to make every AI dev rewrite their roadmap. GPT-6.1 Sol, Dots, Ultrafast, cloud agents, Decisions API, a $500 plan, Codex upgrades... Here’s what actually matters 👇
101
Melvynx @melvynx.dev · 29/09/2026
Unpopular opinion: Sonnet 5.5 should not have been released. In all my tests, it is far worse than Opus 5.5 in terms of results and more expensive. Basically, compared to Opus, it's completely useless. Use Opus 5.5 medium/high/x-high, but not Sonnet 5.5. Sorry.
000
Melvynx @melvynx.dev · 28/09/2026
it's like 20 minutes i want for this shit to be done i ask opus 5.5 to clean my mac and he put half of my mac in the trash
000
Melvynx @melvynx.dev · 26/09/2026
i dream or people is starting to understand the value of lumail.io ?
000
Melvynx @melvynx.dev · 24/09/2026
Life of having a email services business: phishing is everywhere! They come, pay $20, try to send phishing emails, and my AI check system bans them directly It's fast, it brings me money, but not recurring money If you are a serious business, come to lumail.io
000
Melvynx @melvynx.dev · 24/09/2026
Wow, my little Lumail.io SaaS in the most competitive industry reached $480 MRR and a +1,100% growth 🤯 This is insane after 2 years of work and more than 4,500,000 emails sent! This is just the beginning; this app is amazing 🚀
110
Melvynx @melvynx.dev · 11/09/2026
app agent-burn.melvynx.dev
agent-burn.melvynx.dev
Agent Burn — Your AI usage, in view.
A native macOS app and open-source CLI for Codex, Claude, Cursor and more. Track spend, subscription limits and daily usage history.
010
Melvynx @melvynx.dev · 11/09/2026
i try hard to keep the limit positive :
100
Melvynx @melvynx.dev · 09/09/2026
People complaining because Astra uses "too much" weekly limits: don't use it for basic tasks. People ask Astra to create a to-do list; use Sol or Terra for this. Use Astra for complex tasks or creativity only. I am on 5.6 Sol medium most of the time.
000
Melvynx @melvynx.dev · 09/09/2026
Do you think Grok 4.7 will be > Astra ?
000
Melvynx @melvynx.dev · 07/09/2026
comment any better model than this one actually :
000
Melvynx @melvynx.dev · 05/09/2026
Get some fun this week to update the promise of Lumail.io It's basically the AI tool that will finally make you love email marketing Stop feeling like it's a nightmare Just use Lumail and everything goes smoothly
030
Melvynx @melvynx.dev · 05/09/2026
What is crazy about Astra is the quality of output tokens. In deep speech, Astra uses four times fewer tokens than Opus 5 and still performs "better."
000
Melvynx @melvynx.dev · 04/09/2026
chat gpt 0 : melvyn 1
000
Melvynx @melvynx.dev · 04/09/2026
I created my own “Code OS” to manage all my dev servers on my VPS: I can track RAM usage, monitor running apps, kill apps, and more—all through a web UI 🚀💻
000
Melvynx @melvynx.dev · 04/09/2026
how do you feel when you still have one project on master ?
000
Melvynx @melvynx.dev · 02/09/2026
On the $100 plan : 1 agent running for 14 minutes in Fable 5.1 already 10% of session for $100 you can basically do NOTHING with Fable 5.1 lol
010
Melvynx @melvynx.dev · 02/09/2026
I need to resubscribe to test Fable 5.1 🥺 It's crazy that we still have Fable limits, knowing that it's the only reason we pay this subscription; we can only use it 50%.
000
Melvynx @melvynx.dev · 01/09/2026
how to know that we are fucked ? i spend $16k in 1 month if we compare with api price and i only pay $200
000
Melvynx @melvynx.dev · 01/09/2026
ok if @thsottiaux is nice, this time i can say that a reset feel good after i destroy my weekly limits in one day lol
000
Melvynx @melvynx.dev · 31/08/2026
Why do you need to run your AI Agents in a VPS ? Bro this fucking Harness use all the CPU it's the only solution lol
010
Melvynx @melvynx.dev · 31/08/2026
Coding feature of Monday : adding Survey on Lumail.io In SO simple way, fully agentic track the answer, create it, and everything with your favorite agents on Lumail
010
Melvynx @melvynx.dev · 31/08/2026
Every good developer I know thinks that Opus 5 is one of the worst frontier models to code. Fable 5 is still smart, but the Cursor + Codex combo is enough to completely skip Claude. I don't miss Claude at all.
010
Melvynx @melvynx.dev · 28/08/2026
50 views and 19 € the "member first" video generates revenue even before the video generates views it's really nice to be supported by incredible people on youtube
000
Melvynx @melvynx.dev · 28/08/2026
time to switch OFF Hetzner after the completly block my IP expensive and useless i am migrating to netcup again
000
Melvynx @melvynx.dev · 27/08/2026
I just set up a VPS exclusively for coding. No more code on my Mac. Everything is done remotely through the VPS.
000
Melvynx @melvynx.dev · 25/08/2026
the best agent newseltter service is now 100% agentic friendly with everything so your agents one-shot your marketing on lumail.io
000
Melvynx @melvynx.dev · 25/08/2026
It's sad that Lumail.io doesn't have enough MRR to be the #1 place in Outbid. It would be a good investment, I guess, with all the builders that don't use email marketing right for their product. But yeah, next year I will be top #1 in Outbid!!!!
010
Melvynx @melvynx.dev · 25/08/2026
I have an MRR of $200 a month with my SaaS. ... Ah no, it's just my YouTube channel 😂 Thanks for all the support on YouTube. You are the GOAT, especially since I made over 500% since the "drama"!
010
Melvynx @melvynx.dev · 21/08/2026
One change makes my Mac so much healthier: Link cursor and Codex to my VPS named SteveClaw Run all my coding tasks on the VPS Merge with no local test; let the agent handle everything My Mac can finally breathe. It's literally the best move ever
000
Melvynx @melvynx.dev · 21/08/2026
Crazy number : 4 millions 🤯 e-mail sent using Lumail.io ! And each e-mail bring $$$$ so if you have SaaS, I create an article about strategy you can do to make more revenue with e-mail marketing 💸 x.com/melvynx/sta...
000
Melvynx @melvynx.dev · 21/08/2026
i am happy to pay this i am unhappy to pay Claude
000
Melvynx @melvynx.dev · 20/08/2026
Started marketing 2 weeks ago and already hit $100 MRR on Lumail 💃 Let's crack it
000