I’ve been running open-weight models locally for a while. GLM-5.2 on a 512GB Mac Studio was capable, but you always felt the ceiling: 2-bit quant, 5-9 tok/s, tool calls that got flaky in long agent sessions. Local was the thing you tolerated to keep data on your network. DeepSeek V4 Flash was the first real step forward — finally a local model with a genuine balance of performance and speed, one you could actually leave running under agent traffic without babysitting it. It proved the “fast MoE with real agentic chops” formula worked. 5.3-Flash on two DGX Sparks takes that further. 320B params but only 18B active, so it’s running at good speeds. 1M context. Thinking is always on but tunable (low/high/max), and at max it holds up in real agent work through Hermes: multi-step coding, sustained tool calls, no hand-holding. I’m still reaching for cloud models. But for the first time I can see a point coming where you won’t have to. The gap is closing faster than I expected — if you have the hardware, this is the one to try.

If you’re on a Cursor plan and lean on GPT models, mark the date. OpenAI announced Friday it’s ending its partnership with Cursor after SpaceX closed its $60B acquisition. Proposed cutoff: November 12. What changes: your Cursor plan’s included usage (Ultra gives you $400 of model credit) will no longer apply to OpenAI models. You’ll likely need to bring your own OpenAI API key and pay separately, or shift that spend to Claude, Gemini, or Grok inside Cursor. What doesn’t: the companies are still negotiating, so this isn’t final yet. Anthropic already said it’s adding Claude compute for Cursor. The bigger picture: this is the second time this year a lab has drawn a hard line on how its models are accessed (Anthropic did it with OpenCode). Model access is becoming a bargaining chip, and where you build is now a strategic decision, not just a tooling preference. If you’re heavy on GPT in Cursor, start testing alternatives now rather than in November.

Everyone says lock in and stay in your room until the software is done. I think that is one of the biggest mistakes you can make. I drove up to San Francisco, had lunch with a friend, and got the kind of note you only get when the app is in someone else's hands. Make the font bigger so people can still read it while they eat. Tonight I am ripping the server out of SayRay so it runs on the phone. Watch: https://youtu.be/s0yAq8af-Fs

Tech debt. After nearly a year of developing my niche, highly complex voice-and-AI-native software, I am declaring a RESET day. I am unsure whether to celebrate or cry. I am archiving my entire codebase and Supabase schema, choosing to rebuild rather than struggle with refactoring the heavy load of tech debt in the voice-first system I began last October following four months of full-time research. I am shifting from Supabase Cloud to Postgres with PITR and WAL-G, relocating self-hosted agents and communication to memory-optimized Azure Virtual Machines for the LiveKit real-time voice streaming cluster and agent engines. I am archiving the 1.5 million lines of code for a simpler structure, likely using Tailwind CSS with Next.js, React, and Tanstack Query. I am also moving Vercel hosting, Resend, and Twilio to a Canadian Azure VPS and ACS. I chose Azure because they supported my startup with a generous chance to accelerate without bankruptcy :) and without requiring VC funding for higher support tiers, unlike the other major players (I am looking at you, @googledevs and @awscloud) - if you change your mind, reach out! ;) **So tonight, it is a fresh start.** ——— Thanks to @Ray Fernando and community - that last Builder Room conversation got me thinking… and then, seeing how my agents have evolved sooo much in two months, it feels like the right time to start fresh, and ship fast!

Cursor published the real Grok Bot billing page. Read this before you buy a second plan. https://cursor.com/help/grok-bot/plans You do not need a separate Grok Bot subscription. Cursor Pro, Pro+, Ultra, or a self-serve Teams seat already includes it. Sign in with the same Cursor account. Teams does not need a Premium seat. Usage is weekly and it lives on your Cursor account, not on grok.com. Mac and iPhone share one bucket. When the weekly included amount is gone, extra work can keep going on shared on-demand if you have that on. SuperGrok does not pay for on-demand. SuperGrok / Plus / Heavy (individual only) can link as a usage grant. That is not a Cursor plan. It does not cancel Pro/Pro+/Ultra/Teams. SuperGrok Team, Enterprise, and Lite do not link. The link is permanent. You cannot move it yourself, so sign in with the right Cursor account first. The free trial is a usage credit (plus a 7-day window), not a set number of chats. One long agent run can burn the whole thing. Used credit is not restored. If you were in the P-Stack surprise-bill thread: Grok Bot included usage is separate from Cursor “Other Models.” Named Claude/GPT/Gemini still hit that other bar. Same login, same on-demand wallet. Questions, drop them here. I’ll point at the doc instead of guessing.