OpenAI Luna Model Costs 20x More, Online Business Market Thrives

OpenAI's cheapest new model is 20x more expensive than previous generation, with the new Luna model costing $1/M input tokens and $6/M output vs. $0.05/$0.4 for older models, impacting AI startups and developers. — via 1
Multiple profitable online businesses are listed for sale on Flippa and acquire.com, including a car e-commerce platform with $565k annual net profit, an AI newsletter with $265k TTM revenue, and a logistics marketplace with $228.2k TTM revenue, indicating a strong market for digital assets. — via 1 2 3 4
New tools are emerging: Alex Lieberman launched Multi for collaborative website editing, @levelsio built Pietflare AI-powered DDoS detection, Nick Huber released a free tax calculator, and Tony Dinh is developing an AI app for non-technical business owners. — via 1 2 3 4
The "full-stack creator" role is gaining traction, with Sam Parr and Alex Lieberman both emphasizing trust distribution as a moat and hiring 10-20 such creators. — via 1 2
Noam Kagan canceled Evernote after price increase, using Claude to export to Apple Notes, and noted that SaaS pricing must justify value. — via 1

OpenAI's latest model lineup lacks nano and mini options; the smallest new model "Luna" costs $1/M input tokens and $6/M output, while the previous nano was $0.2/$1.25 and older generation $0.05/$0.4, representing a 20x increase in base compute cost. This shift could drive startups to alternative models or more efficient fine-tuning. — via 1
@levelsio found that Claude Code is blocked by content filtering when handling country code/country name tasks; the issue has been reported but not yet fixed, highlighting potential limitations in AI code assistants. — via 1

Flippa lists a profitable car e-commerce platform with $1.7M revenue, $565k annual net profit, 33% margin, including $1.5M inventory and equipment, rated 5 stars. This signals strong demand for established e-commerce assets. — via 1
acquire.com has multiple new listings: an AI news and intelligence newsletter ($265k TTM revenue, $375k asking), a logistics marketplace ($228.2k TTM, $890k asking), an AI image generation SaaS ($106.2k TTM, $200k asking), and a custom LMS with $83k ARR and zero churn ($75k asking). These listings show a diverse range of online business opportunities with clear valuation metrics. — via 1 2 3 4

Alex Lieberman launched Multi, a tool that makes any website collaboratively editable like Google Docs, with comments and AI-powered question answering. It aims to solve feedback loops in document workflows. — via 1
@levelsio built Pietflare, a DDoS and probe detection system that uses AI to analyze access logs and automatically block malicious IPs, ASNs, and countries, with a central control panel. It's a custom security solution for indie developers. — via 1
Nick Huber released a free interactive calculator for cost segregation and bonus depreciation, allowing property owners to estimate tax savings by inputting property type, depreciation basis, and service date. — via 1
Tony Dinh aims to build a simple AI app that lets non-technical business owners fully manage online presence including website building, geo-SEO, social media updates, and Google Maps/Business management. This targets a gap in small business automation. — via 1

Sam Parr and Alex Lieberman both highlight the rise of "full-stack creators" who leverage trust distribution as a moat in the AI era; they are hiring 10-20 such creators at zero marketing cost, focusing on high-quality content. — via 1 2
Noah Kagan canceled Evernote after a price hike, used Claude to export content to Apple Notes in 3 clicks, and argued that SaaS subscriptions must deliver clear value or users will churn. — via 1
Scott Galloway shared a formula for economic safety: annual spending × 25 (assuming 4% after-tax return). His personal figure of $1.25B highlights how lifestyle inflation (children, NYC, greed) drastically increases the target. — via 1

Discover