Neel Techie @TheTechieN

Everyone thinks you need to be in SF to do deals But I'm a 20-something founder running a micro PE fund, based out of Mumbai and buying businesses off the internet for $100k to $1M. So it turns out the deals don't check your zip code.

Wednesday! what are you working on? share your startup / product / saas / whatever you are proud of

Aubergine, an agency theme built with Astro v7 & Tailwind CSS v4. ✓ 10 ready-made pages ✓ Cursor rules & AGENTS files included ✓ Content Collections ✓ SEO optimized ✓ Astro + Sanity CMS v6 available ✓ Lifetime updates & support ✓ GitHub access with bundle Take a look: lexingtonthemes.com/templates/aube…

keyboard_arrow_left Previous keyboard_arrow_right Next

decent standard-format launch video - amazing visual hook. the blueberry hat was a great choice - good camera quality - decent font choice - very dialed in script - good CTA but - for demo would say "let's say you're a ICP and you PROBLEM" - blue text a little hard to read on the background - very generic launch video format im tired of - animations for the initial demo was a bad move because it feels faked. would have had a real UI - music is very generic and lame. also a little quiet. - "already made millions" - who did it make millions? - would have had customers in the video this is well executed, but its the format I'm growing very tired of. something they did that was unique was the visual hook. think more founders should think about random things like this hat to include in the first 5 seconds to keep people's attention. again, solid execution but can't keep giving this format top scores any more because its so overplayed

Sean Rich @TheSeanRich

22 hours ago

Introducing Blueberry: the first commerce agent that sells for you. You give it a goal. Blueberry finds everyone who engages with you, researches their profiles, and sends the right DMs at the right time to drive sales. It's already driven millions for brands. See how 👇

103 267 2K 1.8M 2K

I always assume launches won't move the needle much But yet, a voice in my head said "what if it works" I'm very thankful I listened to that voice and took a few days to setup K8 with proper autoscaling rules Volume on API this week has been absurd

Build apps with any AI Chat or Agent. Deploy them to a live URL. Then share them securely with your friends, family, and co-workers. Works with ChatGPT, Claude, OpenClaw, and many more.

🛠️ Jean 0.1.54 - 2 weeks of improvements, features, and fixes. - Mr. Robot (beta) - auto-investigate and fix issues regularly, even set off hours, so it can do it while you sleeping. - Command Code CLI (beta) support - PI CLI (beta) support - WSL support - Steer support for Codex, Opencode, PI, so you can add new thoughts in the middle of the prompt. - Change AI providers between prompts. And bunch more.

keyboard_arrow_left Previous keyboard_arrow_right Next

this might save you money on apple search ads, give it a try. i built a small module that tries to find the "right bid keywords" for an app. (no, not llm) each query costs ~2$ (not llm cost), so i need real volunteers to understand if this really works and how to improve it. if you’re running ads and want to see what keywords you should be bidding on (not just obvious ones), drop: your app link (us for now) 2 seed keywords i'll run it and send results back looking for a few volunteers to validate this in the wild i tested with Tinder for US (not affiliated, just picked randomly), you can see the results below.

not many things make me angry but someone telling me that i have to do something a certain way, or telling me how to live my life and make my decisions that does get to me a little bit

this focus app gets 20k downloads/month from a chicken 😭 if you leave the app while focusing... the chicken dies 💀 onboarding: - explains the concept instantly - choose how strict you want focus mode - builds emotional attachment immediately - shows stats and asks for notifications - paywall shown after users understand the value - includes a lifetime option most productivity apps sell discipline this app sells guilt such a simple concept, but brilliant psychology build your own app with Anything

Anthropic just showed how to get frontier-model quality at 5x lower cost without running a frontier model Here's how their platform team laid it out on stage: A cheap model drives. It only "phones" Opus when it's stuck +2.7 pts on SWE-bench, ~12% cheaper per task Watch it ↓

Alex @de1lymoon

3 days ago

The creator of Claude Code wrote 0 lines of code by hand in 2026 His record: 150 PRs shipped in a single day from his phone 100% of his code is now written by Claude Code In ~25 min he maps what comes next: loops, agent orchestration, and why Claude Code may shrink to ~100

15 3 43 5K 36

i can’t believe we can vibe code multiplayer games now. made with higgsfield + claude fable 5. what are we going to build in 1 year?

Higgsfield AI 🧩 @higgsfield

6 days ago

Meet Higgsfield Games. For the first time, build and deploy multiplayer games from one prompt, in any genre, 2D or 3D, with best-in-class characters, props, and settings generated by Higgsfield MCP. Powered by Claude Fable 5. Try on Claude via MCP and on our Supercomputer.

291 332 2K 900K 1K

Based on a few recent posts from people comparing their $12k usage limits maxxing. This can now jump up to $ 24k???! Holy moly. I'm prepping my loops as we speak

OpenAI @OpenAI

6 days ago

We heard you wanted to use Codex rate limit resets on your own time. Starting today, we’re rolling out the ability to save rate limit resets to use later. We’re starting Go, Plus, Pro, and Business users with one free reset:

1K 2K 22K 4.4M 4K

ZeroIndex An documentation theme for Astro & Tailwind CSS. It's also available in a variant for Sanity CMS. Take a look: lexingtonthemes.com/templates/zero…

keyboard_arrow_left Previous keyboard_arrow_right Next

Patient to agent is the future of healthcare

Christina Farr @chrissyfarr

5 days ago

This is actually insane. The patient facing tools are performing better than the clinician facing tools. Direct model access is a big deal. 🤯

16 19 145 12K 54

THIS GUY USED A $20 CLAUDE FABLE 5 SUBSCRIPTION TO BUILD A MACOS-STYLE WEB OS WITH 1 PROMPT AND 1 HTML FILE one page with embedded css + javascript now has a menu bar, dock, 3 draggable windows, desktop icons, notifications and a working app layout inside the browser this is not a figma mockup and not a frontend team spending 14 days rebuilding macos for a demo, this is a single generated html file acting like a small operating system the wild part is the feature count. file explorer, terminal, browser mockup, settings app, audio player, boot screen and a playable minecraft-style clone all running in the same fake desktop studios can charge $3,000 to $7,500 for interactive web demos because clients are buying something they can click, test and understand in 60 seconds the breakdown below covers how this same workflow turns prompts into sellable mini-games: 6 prompt layers, one-minute deploy, $20 build cost and $500 to $2,000 client reskins

Gipp 🦅 @gippp69

6 days ago

x.com/i/article/2064…

23 6 111 168K 121

Kimi just dropped K2.7-Code > open sourced > 21.8% better on Kimi Code Bench v2 > 11.0% better on Program Bench > 31.5% better on MLS Bench Lite > uses 30% fewer reasoning tokens > follows instructions more accurately > better at long-horizon coding tasks > higher end-to-end coding success rates open-source coding models keep improving faster than most people expected.

Kimi.ai @Kimi_Moonshot

5 days ago

🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced! 🔷 Improved coding & agent performance over K2.6: +21.8% on Kimi Code Bench v2, +11.0% on Program Bench, and +31.5% on MLS Bench Lite. 🔷 Reasoning efficiency: Less overthinking, with 30% lower

keyboard_arrow_left Previous keyboard_arrow_right Next

630 2K 14K 2.4M 3K

I think everyone should build a personal brand on X. You can get tons of credits from great products, use the latest models, and share your journey while going full buildmaxxing mode. 🫡

keyboard_arrow_left Previous keyboard_arrow_right Next

Today. Live stream link for those who want to follow along live: youtube.com/live/ZTOFs1Uk_…

ROSS @ROSSIntel

2 months ago

The Third Circuit Court of Appeals has scheduled oral argument in Thomson Reuters v. ROSS Intelligence for June 11, 2026.

3 3 21 14K 6

Surprise, surprise! Turns out the hard part was never the code. It was knowing what to build and designing it so people stay. Most valuable skill.