Posts
Elon Musk
@elonmusk · 9 juil. 2026
Yes
Yun-Ta Tsai@yunta_tsai· 9 juil. 2026The road to AGI is to teach it to solve Kardashev Type II problems.
523
235
1,6 k
721,9 k
Elon Musk a reposté
NVIDIA
@nvidia · 8 juil. 2026
Congrats to @SpaceXAI on Grok 4.5 — trained on NVIDIA GB300 NVL72 systems and purpose-built for coding, agentic tasks, and knowledge work. This is what happens when world-class AI infrastructure meets world-class AI research.
SpaceXAI@SpaceXAI· 8 juil. 2026Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.
148
354
3 k
277,8 k
Elon Musk a reposté
Sawyer Merritt
@SawyerMerritt · 8 juil. 2026
Smart pet technology company Fi has just launched the world's first dog collar with @Starlink.
Basically, you’ll know exactly where your dog is at all times anywhere in the U.S. It’s $199 + $189/year membership for new customers, and just a $299 flat fee for existing members.
More info:
• Up to 5 days of battery life: Operate multiple days on a single charge, supporting extended trips and remote exploration.
• T-Satellite with Starlink GPS: Track a dog's location nearly anywhere in the US, including areas entirely without cellular coverage.
• Dual Band High-accuracy GPS: Deliver precise location data continuously, whether connected via satellite, LTE, Wi-Fi, or Bluetooth.
• Automatic switching: Seamlessly connect to LTE when available and to T-Satellite with Starlink when cellular networks are out of reach — with no action required from the owner.
• Real-time location updates: Deliver accurate, continuous location data whether the dog is in a city park or miles from the nearest cell tower.
• Fi Callback: A proprietary training system based on sound and vibration that uses no static electric shock to bring dogs back to their owners on command.
• Fully waterproof: Tested in salt water and IP68/66K-rated
280
400
4,3 k
460,4 k
Elon Musk a reposté
Déborah
@dvorahfr · 8 juil. 2026
Wow amazing Grok Imagine 😍
134
95
620
120,4 k
Elon Musk a reposté
DogeDesigner
@cb_doge · 8 juil. 2026
BREAKING: SpaceX’s Starship just got another lunar customer. Japan’s ispace will use SpaceX’s Starship for Moon cargo missions.
• ispace signed an agreement with SpaceX to secure 500 kg of payload capacity on Starship.
• First mission is targeted to launch as early as 2030 through ispace’s new Lunar Access Integration service.
• The service is designed to make it easier for companies, governments, and research organizations to send payloads to the Moon.
• Customers can fly payloads ranging from a few kilograms up to 500 kg on shared Starship missions.
“We are very pleased to be able to offer the new Lunar Access Integration service utilizing Starship’s payload space… High-capacity, relatively low-cost lunar transport… is essential to realizing the sustainable lunar economy that ispace aims to create.”
— ispace Founder & CEO Takeshi Hakamada
202
553
2,5 k
169,6 k
Elon Musk
@elonmusk · 9 juil. 2026
Neat
Box@Box· 9 juil. 2026Grok 4.5 reviewed a full Credit & Security Agreement stored in Box — the kind of dense, multi-section facility document that typically requires significant counsel time.
@Grok 4.5 used Box MCP to access the file securely, extract key terms across the agreement, identify
468
200
1,1 k
639,8 k
Elon Musk a reposté
Dan
@Daniel_Farinax · 8 juil. 2026
Grok 4.5 is actually insane.
I'm one step closer to launching my new 3D Home Plan Builder. Powered by Grok 4.5 and Grok Imagine. 🚀
Grok 4.5 just crushed it: designed the entire 2D + 3D plan inside the app in under 1 minute at a fraction of the cost of Opus.
Launching tomorrow! 🔥
105
157
1,2 k
142,8 k
Elon Musk
@elonmusk · 9 juil. 2026
Cool that Grok 4.5 is #1 in some respects, even with respect to Fable 5
Artificial Analysis@ArtificialAnlys· 9 juil. 2026SpaceXAI's Grok 4.5 takes the #1 spot on AutomationBench-AA with a score of 51%, ahead of Claude Fable 5 (49%) and Claude Opus 4.8 (48%) at roughly a quarter of their cost per task - the first model to complete more than half of workflow objectives without breaking any business
640
302
2,1 k
729,4 k
Elon Musk a reposté
Teslaconomics
@Teslaconomics · 8 juil. 2026
Wow… within 5 minutes my kids built this game called the Tesla City with Grok 4.5.
It’s a simple Tesla arcade game in SF where:
1/ You can choose a Tesla (3 / Y / S / X / Cybertruck / Roadster / Semi / Cybercab)
2/ Give robotaxi rides - pickup and drop people off for $
3/ Can upgrade your car with the $ (buy new weapons, etc.)
4/ Fight gas cars with blasters, homing missiles, EMP & plasma, as they try to attack you to prevent you from completing Robotaxi rides.
GROK 4.5 is next level! 🔥
Teslaconomics@Teslaconomics· 8 juil. 2026My 5 and 7 yr old are ready to cook with Grok 4.5!
It’s going to be a fun summer…
134
129
825
225,6 k
Mark Zuckerberg
@finkd · 9 juil. 2026
(1) Today we're releasing Muse Spark 1.1 -- a strong agentic and coding model at a very low price. It's available through our new Meta Model API and in Meta AI.
Elon Musk
@elonmusk · 9 juil. 2026
Jinx
890
473
17,8 k
936 k
Drew Pavlou 🇦🇺🇺🇸🇺🇦🇹🇼
@DrewPavlou · 8 juil. 2026
Grok 4.5 delivering Fable level performance at like 1/17th the cost! 😮 @elonmusk
Elon Musk
@elonmusk · 8 juil. 2026
In fairness, Fable is definitely better than Grok 4.5, but most tasks don’t require Fable-level capability
803
628
20,7 k
941,7 k
Elon Musk
@elonmusk · 8 juil. 2026
Grok has always been very strong on law
Tech Dev Notes@techdevnotes· 8 juil. 2026Grok 4.5 is #1 on Harvey's Legal Agent Benchmark
711
415
2,6 k
819 k
Elon Musk a reposté
DogeDesigner
@cb_doge · 8 juil. 2026
BREAKING: Tesla Cybertruck’s bulletproof body helped protect a toddler, an infant, and neighbors during a shooting.
Cybertruck became a real-life shelter during a chaotic 4th of July neighborhood party.
• Owner says guns were drawn after an altercation
• Cybertruck sheltered a toddler, an infant & neighbors until police arrived
• Earlier, the truck was even powering the block party
• “You hope you never need that kind of protection, but in that moment, I was very glad it was there.”
Source: FB/Cybertruck Owners Only
272
668
3,6 k
243 k
Elon Musk
@elonmusk · 8 juil. 2026
Rate of improvement is 🚀🚀
Aman Sanger@amanrsanger· 8 juil. 2026Very excited for this model. It’s an enormous improvement over composer 2.5 and trained entirely from scratch. It’s been a pleasure working with the SpaceXAI team on it.
Even more excited by the slope of the effort and the models to come.
436
283
1,8 k
723,7 k
Elon Musk
@elonmusk · 8 juil. 2026
Rate of improvement is accelerating.
Users should notice a meaningful improvement in the usefulness of the Grok Build harness with our V9 foundation model (aka Grok 4.5) every week.
Gavin Baker@GavinSBaker· 8 juil. 2026Grok 4.5.
Pareto dominant for coding by the numbers.
We will see on the all-important vibes.
Instinct is the benchmarks are likely directionally accurate given the stated focus on real world utility.
458
331
2,1 k
764,4 k
Elon Musk
@elonmusk · 8 juil. 2026
Grok 4.5 is not yet using our internally developed C/C++ inference software that exact maps to the GB300 hardware. Doubling or more of the current speed is probably achievable.
Farzad 🇺🇸 🇮🇷@farzyness· 8 juil. 2026I think AI has just hit a gigantic threshold, and Grok 4.5 is the PERFECT example as to why that is.
One of the hardest parts of working with AI is iterating on a project or task that you're working on.
As the models have gotten smarter (and more expensive), it's taking longer
500
396
3,4 k
836,9 k
Elon Musk a reposté
Andrew Milich
@milichab · 8 juil. 2026
Grok 4.5 brings frontier performance across coding and knowledge work
Artificial Analysis@ArtificialAnlys· 8 juil. 2026SpaceXAI’s Grok 4.5 scores 54 to place fourth on the Artificial Analysis Intelligence Index following only Fable 5, GPT-5.5, and Opus 4.8. It scores on par with GPT-5.5 in Codex on the Artificial Analysis Coding Agent Index in the Grok Build harness, at much lower cost
Grok 4.5
73
136
703
235,3 k
Elon Musk a reposté
Artificial Analysis
@ArtificialAnlys · 8 juil. 2026
Grok 4.5 in Grok Build also stands out for its efficiency. Grok 4.5 in Grok Build cost $2.49 per task while Fable 5 in Claude Code cost $11.80 and GPT-5.5 in Codex $5.07. This is driven by relatively low token pricing and the model using far fewer tokens than comparable models (1.9M average tokens used per task), significantly less than Fable 5 in Claude Code (7.2M) and GPT-5.5 in Codex (6.2M)
72
147
652
161,5 k
Elon Musk
@elonmusk · 8 juil. 2026
Grok 4.5 true usefulness is excellent
Snorkel AI@SnorkelAI· 8 juil. 2026We obtained early access to evaluate @SpaceXAI's newest Grok 4.5 model on GDPval+: real professional deliverables (from a variety of economic sectors and occupations), graded against expert-authored rubrics. Grok 4.5: 29%. GPT 5.5: 22%. Claude Opus 4.8: 21%.
537
392
2,1 k
887,9 k
Elon Musk
@elonmusk · 8 juil. 2026
Grok 4.5 context window will upgrade to 1M probably by next week
Artificial Analysis@ArtificialAnlys· 8 juil. 2026SpaceXAI’s Grok 4.5 scores 54 to place fourth on the Artificial Analysis Intelligence Index following only Fable 5, GPT-5.5, and Opus 4.8. It scores on par with GPT-5.5 in Codex on the Artificial Analysis Coding Agent Index in the Grok Build harness, at much lower cost
Grok 4.5
426
364
2,3 k
848,8 k