Hey friends, hope your week is off to a great start!
I've been digging through a ton of new tools and tech news over the past few days, and honestly, things are moving crazy fast.
Today, we're talking about how AI is completely changing how SaaS companies charge us (say goodbye to flat per-seat pricing!), a brand new open-source note-taking space that lets coding agents write right alongside you, and a major new model drop from Anthropic that is going to make running agent loops a whole lot cheaper.
Finally, before we get into it, I just wanted to mention that we’ve crossed our waitlist of 150 members. Huge milestone for us. Over the next few weeks, we’ll be adding more features and tools to our platform. Our goal is to make Workstak the one-stop portal for understanding SaaS tools and the future of pricing, and to give you a better sense of choosing the right products for your complex AI workflows. Thanks for joining.
Grab a coffee (or tea) and let's dive in…
📰 The Latest Happenings:
Claude Sonnet 5 is here: Anthropic just dropped Claude Sonnet 5, which is a massive deal for developers running AI coding agents. They've optimized the model for raw speed and cheap tool use, making it ideal for running multi-step agent loops, fast background chores, and instant code autocomplete without burning through a budget. While the Opus tier (like Claude 4.8) remains the powerhouse for deep reasoning and complex document parsing, it comes with longer response delays. Anthropic is clearly splitting these lines: Sonnet 5 is the speedy workhorse for daily developer loops, while Opus handles heavy-duty, high-accuracy reasoning.
Cursor launches a mobile App: Following its recent acquisition by SpaceX, Anysphere (the team behind Cursor) has launched a mobile companion app for iOS and iPadOS. Instead of trying to edit code on a small screen, the app acts as a supervisor dashboard that lets developers monitor and guide autonomous coding agents remotely. Users can approve file writes, review pull requests, and prompt coding agents through refactoring tasks. This shifts coding from a desk-bound activity into a flexible, supervisor-style mobile workflow, allowing teams to keep continuous deployment cycles running smoothly during commutes or away from their workstations.
AI demand sparks global RAM shortage: A major boom in AI servers is causing a global RAM shortage and driving up consumer hardware prices. Chip giants like Samsung, SK Hynix, and Micron are reallocating production lines away from standard consumer DRAM to supply lucrative high-bandwidth memory (HBM) to data centers run by OpenAI, Google, and Meta. This reallocation is pushing up device pricing for laptops, PCs, and gaming consoles (including Apple Mac, Xbox, and Nintendo devices), with IDC predicting the shortage could last until 2027. For developers looking to build local configurations or run resource-heavy IDEs, it means paying premium component rates; shifting to optimized apps or remote cloud environments can help bypass the need for expensive local upgrades.
Google Gemini unlocks free image editing: Google has made its personalized AI image generation and editing features in Gemini free for all users in the U.S. Powered by its newest Imagen 3 model, the update lets users generate custom graphics, edit specific image regions (inpainting), and maintain character consistency across multiple frames. This gives builders and designers a free way to prototype mockups and create marketing assets, stepping up Google's competition with paid services like Midjourney and DALL-E 3.
OpenAI Previews GPT-5.6 Sol: OpenAI has released a limited preview of its new GPT-5.6 model series, led by the flagship reasoning model, Sol. Due to U.S. government export controls and safety guidelines, access is highly restricted, causing deployment delays for international developers. Sol is designed to compete directly with Claude Opus and Mythos on SWE-bench coding and terminal tasks while using fewer tokens. OpenAI has priced Sol at $5.00 input / $30.00 output per million tokens, sitting alongside the mid-tier Terra and budget Luna options.
🔥 Interesting Finds:
AI agents are taking over traditional employee roles, and it's starting a massive shift in how SaaS companies bill us. Since seat counts are shrinking but software use is growing, over seventy percent of vendors are moving to usage-based or outcome-based contracts (like HubSpot charging per resolved chat). It's great because we pay for real outcomes, but it also means we'll need to keep a close eye on our credit balances so we don't get hit with massive surprise bills.
If you like Notion or Obsidian but want something that plays nice with AI, check out OpenKnowledge. It's a local-first markdown editor built so that desktop agents like Claude or Cursor can open up the editor and write side-by-side with you. It uses Git to sync changes and keep your data private, which makes it super easy to build a collaborative writing loop with your AI assistant.
As coding tools standardize on the Model Context Protocol (MCP), deploying and hosting custom servers can get annoying. Manufact wants to be the "Vercel for MCP." They connect directly to your GitHub repo to deploy servers with preview links, and they include automated testing tools to help you verify everything is running smoothly.
A developer put eighteen years of Hacker News comments and posts into a public, real-time ClickHouse database so you can query tech trends for free. You can run simple SQL queries to compare buzzwords over time (like watching "blockchain" slide while "OpenAI" takes off), making it a goldmine for anyone looking to analyze community interest without having to write a scraper.
Here are some more screenshots:
🌐 Across the Socials:
Ara is the lead economist at Ramp, and he shared some findings analyzing over 21,000 U.S. firms. The data shows that companies adopting AI at high intensity actually grew their headcount by about 10% over the two years following adoption. More surprisingly, entry-level hiring rose by 12% across multiple departments, including support, sales, and engineering, challenging the fear that AI is purely a job-killer.
When you integrate AI into a company, you make the existing team way more productive, which accelerates company growth. And when a business starts growing fast, they naturally need to bring on more people to support that expansion. It's a reminder that AI is a tool for growth and productivity, not just cost-cutting.
Interesting thread showing how Uber is rolling out agentic AI across their entire organization. Looks like 99% of Uber engineers use AI tools, and over 70% of their pull requests are now driven by AI agents. To expand this success to non-technical teams like finance, legal, and marketing, Uber formed 'Agentic Pods'—two-week sprints pairing top AI engineers with domain experts. This led to efficiency gains with capital allocation times falling from 15 hours to just 30 minutes, and financial pacing reports dropping from two days to 10 minutes.
The big lesson here is that the massive wins don't come from just automating a tiny task here and there. You have to step back and rethink the entire workflow from scratch.
Totally. It’s crazy how much knowledge we can tap into right now. The only thing holding us back is our own curiosity. This doesn’t mean you’re an overnight expert, but it means you can at least get a sense of different fields in a much shorter time period.
The narrative claims open source is eating enterprise AI as the capability gap shrinks and adoption grows. Yet open source fell to 11% of enterprise LLM spend from 19% last year. New immature use cases drive frontier closed models while mature ones shift to specialized open weights. So basically, open source will own production as use cases mature, but share won't increase for many years.
🚀 How you can support:
Thanks so much for spending time with us this week!
We love building this community, and if you want to support what we're doing here, there are a few quick ways you can help out:
Share the love: If you got value from this edition, please forward it to a friend, co-worker, or builder in your network who might find it useful.
Check out our catalogue: We are adding new vetted tools to Workstak's curated tool directory every single week. Go check it out and see if there is something that can level up your workflow.
Snag exclusive deals: If you want to get your hands on exclusive deals for AI tools, go ahead and create an account and let us know what you are interested in so we can hook you up.
Read the blog: Head over to the Workstak Blog for deeper plays, guides, and thoughts on both building and buying software.
Partner with us: Are you building a tool that would be a good fit? We'd love to help you get it in front of the right people. Swing by our partner page and apply to launch with us!
