SpaceX completes record $60B Cursor acquisition

Share
SpaceX completes record $60B Cursor acquisition

SpaceX just closed the largest startup acquisition on record, folding AI coding tool Cursor into a new SpaceXAI division with the brand set to fade into Grok — and co-packaged optics officially hit the mass-production phase of the AI buildout.

SpaceX finalized its $60 billion all-stock acquisition of Anysphere, the company behind Cursor, confirmed via an SEC Form 8-K — the largest startup acquisition on record. The deal issued roughly 391 million SpaceX Class A shares, with Cursor's outstanding shares converting into about 389.3 million of them and additional restricted stock units and options assumed. Cursor now operates as a wholly owned subsidiary under the newly rebranded SpaceXAI division, its team gets access to SpaceX's Colossus supercomputer, and the combined entity has already launched the beta of Grok Bot — the agent app that, as we covered on Tuesday, was the first concrete link between Grok and Cursor's paying subscriber base — SpaceXAI launches Grok Bot, a cross-platform AI agent app.

Analysts expect the Cursor brand to be phased out in favor of the Grok moniker — The Information reported last week that Cursor told staff the brand would likely disappear — and Morgan Stanley, which maintains a $300 price target on SpaceX, projects the acquisition could add up to $13 billion in revenue by 2027. Investors took the closing in stride: shares ticked up 1.5% to $143.45.

The deal completes Musk's play to convert the industry's most trusted coding tool into distribution for his own models — consistent with his claim, covered earlier this week, that AI revenue will top rockets and Starlink by September (Musk: SpaceX AI revenue to top rockets and Starlink by September). The risk is the rebrand itself: developers are a fickle audience, and erasing a brand they trust in favor of Grok hands Anthropic and OpenAI an opening.


Nvidia has officially moved its co-packaged optics (CPO) switches into full mass production, per Chinese business media reporting Friday's announcement. It formalizes a ramp that built through the summer: Nvidia's Spectrum-X CPO switch, developed with TSMC using its COUPE packaging, delivers up to 400 Tb/s of switching capacity, while Broadcom's 51.2T Bailly switch is in volume manufacturing with support from Delta Electronics and Micas Networks — cutting power use by up to 70% versus pluggable transceivers and already validated at hyperscalers like Meta, according to TrendForce.

CPO is the industry's answer to the AI buildout's next wall: data movement and the power it consumes. It's the same optical wave that showed up in Lumentum's doubled quarter (Lumentum's revenue doubles as AI data centers go optical). The bottleneck now is supply — optical engines, silicon photonics and advanced packaging capacity will decide how fast the ramp can go.

What to watch: whether the Cursor-to-Grok rebrand triggers a developer exodus, and how quickly CPO clears its supply bottlenecks.

Would you keep using Cursor after it becomes Grok? Tell us in the comments.

Read more

Byteification turns Qwen3 and Llama 3 into byte-level models

Byteification turns Qwen3 and Llama 3 into byte-level models

The tokenizer is the last piece of the language-model pipeline that was never learned end to end — and a Nature paper out this week shows it can be swapped out for less than one percent of a pretraining budget. Researchers have retrofitted four open subword language models into true byte-level models using less than 1% of a typical pretraining budget, and the result — published 7 October in Nature — closes the performance gap that has kept byte-level models a niche curiosity for nearly a decad

OpenAI's dots go mobile and can start your Codex work

OpenAI's dots go mobile and can start your Codex work

One product update and two benchmark papers this hour, all circling the same question: how much of the work can you actually hand over, and what happens to the part you didn't? OpenAI's dots — the always-on agents that get their own cloud computer — can now be created and configured entirely from the ChatGPT app on iOS and Android, and a dot can start Codex work or pick up an existing Codex thread on your behalf. The October 9 release, billed as the first of a fresh run of updates for dots, mo

Coding agents as a hospital: Cockroach's five-month results

Coding agents as a hospital: Cockroach's five-month results

The agentic-coding debate keeps being framed as speed versus trust. Cockroach Labs just published the most detailed trust-first ledger yet — five months of numbers, including the parts that didn't work. Cockroach Labs ran its codebase like a teaching hospital for five months and says the pipeline merged well over a million lines of code with seven total reverts. In the setup the company calls MOLT Sinai, GitHub issues are patients, merging is discharge, and the humans in charge are Chiefs of M

Today in AI — October 10, 2026

Today in AI — October 10, 2026

A day where the boring machinery took center stage: proof checkers, order books, warehouses and the humble text message — plus fresh arguments about who gets to declare any of it safe. Models & Research * Mathematicians get a reliability primer for the Lean Theorem Prover. A guest post by Thomas Hales on Terence Tao's blog walks through what Lean actually guarantees — and devotes itself to the "Summer of Soundness Bugs," the string of kernel bugs found this summer that let false proofs thr