If you chose anything other than the first option, this edition is for you.

Towards the end of June I was running out of topical ideas for this newsletter.

Very little was moving barring the Fable saga, and I had started sketching a creative batch of editions to fill the gap.

Then July opened the firehose, and I don't think it's closing any time soon.

I'll cover the important ones in dedicated editions over the coming weeks, and if there's a specific one you want first, hit reply and tell me.

This edition the plan is simpler: one wire, everything that happened, so we're all on the same page, grouped by what each one means for you.

Before we dive in:

Partner with us

If today's piece left you wanting a proper handle on AI vocabulary, here is the best resource.

85 terms explained in plain English, structured so each section builds on the last.

Quick "test yourself" checkpoints between sections keep what you learn from evaporating, and 15 minutes of scrolling will leave you a lot more fluent than you were this morning.

Grok 4.5 lands.

SpaceXAI's flagship at $2 in and $6 out per million tokens, trained alongside Cursor. 500k context, and Musk is calling it an Opus-class model.

GPT Live replaces Advanced Voice Mode.

Full-duplex voice that listens while it speaks, so you can interrupt naturally and translate live. The mini model becomes the default for everyone.

SWE-1.7 ships in Devin.

Cognition trained it from a Kimi K2.7 base and serves it via Cerebras at 1,000 tokens per second.

GPT 5.6 goes broad.

Sol, Terra and Luna roll out across ChatGPT, Codex and the API after a limited launch reviewed by the US government. Altman says Sol is 54% more token efficient on agentic coding.

Muse Spark 1.1 opens to developers.

Meta's agentic coding model at $1.25 in and $4.25 out. Zuckerberg posted on X for the first time in three years to announce it.

My take: 4 frontier releases in roughly 24 hours.

I can't think of a precedent for this much cumulative progress landing in one stretch, and every lab is now forcing the others' hand.

The future is absolutely amazing.

My take: Grok 4.5 at $2/$6, and Meta undercutting it at $1.25/$4.25. Then GLM 5.2, open weights under an MIT licence, beating GPT-5.5 on SWE-bench Pro at $4.40 output tokens against Sol's $30, on a chart SpaceXAI published themselves.

For founders, this is a great time to be building. Your model bill is an architecture decision now.

xAI is now SpaceXAI.

The rebrand is complete with a new logo and handle, 5 months after the all-stock merger folded xAI into SpaceX and weeks after the largest IPO in history.

ChatGPT Work arrives, Atlas dies.

OpenAI's workplace agent drafts documents, spreadsheets, decks and web apps, while Chat, Work and Codex merge into a single desktop app. The standalone Atlas browser is being sunset, its agentic browsing features folded into the desktop app and a Chrome extension.

OpenAI calls GPT 5.6 the "preferred model" for Copilot 365.

A pointed claim in the same week Microsoft is publicly routing Copilot traffic to its own models.

Claude Reflect.

Anthropic's beta dashboard shows how you've used Claude over the past year, with quiet hours and break nudges built in. The internet has already renamed it Claude Wrapped.

My take: the interesting fight has moved above the model layer. Who owns the desktop and the enterprise seat now matters as much as who tops the benchmark chart.

My take: image models are shipping on a weekly cadence now. Any leaderboard screenshot has a shelf life of about a fortnight.

Dedicated editions are coming on GPT 5.6 and Grok 4.5, with my own hands-on tests. If there's a model or tool you want covered first, reply to this email and tell me.

And if the rumoured Gemini 3.5 Pro lands next week, this wire will look quaint by next week.

Until then,
Vaibhav 🤝🏻

If you read till here, you might find this interesting

#AD 1

Join beehiiv live on July 16th at 1PM ET for a first look at the future of audience-led business.

This isn’t just another feature launch (though there will be plenty of those). It’s a look at a more connected future for creators and brands that are tired of juggling disconnected tools, platforms, and data.

If you care about building an audience online, this is worth your time.

Reserve your spot.

#AD 2

Every missed onboarding step, delayed offboard, or broken provisioning handoff has a root cause: HR and IT aren't aligned. This guide gives both teams a shared framework for the full employee lifecycle.