William Davis retweetledi
The 10T model is gonna cook
Cursor@cursor_ai
Composer 2.5 is exceptionally intelligent and up to 10x more efficient than similarly capable models.
English
William Davis
654 posts
William Davis
@itswillandbill
Coffee nerd, Future paramedic, tech enthusiast.
Australia Katılım Mayıs 2022
451 Takip Edilen141 Takipçiler
William Davis retweetledi
William Davis retweetledi
Just tap the Grok logo
@jason@Jason
This new Grok menu is 🔥 who came up with this banger? Keep going!
English
Here are the top 7 models on @ArtificialAnlys comparing intelligence score with price per $100 spent for each point. Higher is better. Grok 4.3 is by far the most cost effective model currently. Pushing above its eight for only a 500B parameter model. Great work @xai and @elonmusk
English
William Davis retweetledi
@karankendre Anthropic will not be destroyed. Their AI+harness goes far beyond coding and Opus 4.7 is still better than Composer 2.5, albeit a lot more expensive.
Cursor is however an important piece of the puzzle to make Grok much better.
English
@_navicstein @elonmusk Yeah I’m thinking this more and more
English
William Davis retweetledi
Try Composer 2.5 on Cursor!
Michael Truell@mntruell
Composer 2.5 is now the most-chosen model in Cursor. We're giving everyone 10x usage for the rest of the day. Enjoy!
English
William Davis retweetledi
Starting today, use your Grok or X Premium subscription in @openclaw.
Chat with your agent, generate images and videos, or search for X posts.
x.ai/news/grok-open…
English
Excited to see what @xai cooks up with there v9 1.5T parameter model. With the added @cursor data. With what the data can do to kimi 2.5, composer 2.5 is an incredible achievement and powerful coding model. Grok should take it to the next level. With its high intelligence, agentic abilities and low hallucinations. Grok has a vey good chance to being sota very soon.
@spacexai will acquire cursor, there partnership is just to powerful. I can also see them acquiring @Kimi_Moonshot as well. These together will create some of the most powerful models out there for every task imaginable!
English
William Davis retweetledi
Try it out!
(Partially trained on Colossus 2)
Cursor@cursor_ai
Introducing Composer 2.5, our most powerful model yet. It's more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions. For the next week, we’re doubling the included usage of the model.
English
@elonmusk Any update on grok 4.4, the 1T parameter model?
English
Try it out …
Improvements are landing every few days!
Tesla Owners Silicon Valley@teslaownersSV
Grok Build CLI Beta can now be installed directly from Grok Web with a single terminal command. The agentic coding and workflow tool is currently available exclusively for SuperGrok Heavy subscribers. At the same time, xAI is offering a major discount on the Heavy tier: 67% off for the next 6 months, bringing the price down to $99/month instead of $300/month. xAI is moving fast to get more developers building with Grok.
English
William Davis retweetledi
This is equivalent to trying to sell a cop drugs while he’s in uniform in his police car.
English
With grok build released. We really need to see grok expert and heavy updated within there web/app platforms. Using grok build to research with multiple agents and run tasks autonomously is a game changer and helps utilise the intelligence and skills of the model further. Grok 4.20 multiagent is great and 4.3 is a big step above it. But grok build scaffold and system is on another level
English
@IamEmily2050 In terms of intelligence it already is at opus 4.6 levels. Coding is where we currently fall short. I think we will see the next revision close this gap quite quickly and also build upon 4.3 strong intelligence capabilities
English
William Davis retweetledi
This means the next Grok update is coming at the end of June early July just before the IPO with performance between Opus 4.5 and Sonnet 4.6 not bad and very useful for so many things, but still far from SOTA. Hopefully, they will catch up in the second half of next year.
Elon Musk@elonmusk
We are improving the 0.5T Grok foundation model V8 (public version 4.3) every few days. The 1.5T V9 just finished training (incorrectly called pre-training) and is a major upgrade. Next, we are adding the Cursor data in supplemental training (others call this mid-training), then SFT and RL. About 3 or 4 weeks to release. This will be a banger.
English
William Davis retweetledi
@leonbi100 Congrats. Can’t wait to see what you all cook!
English
It's already such a privilege I get to work on RL for V8.
Holy shit I can't wait for V9.
Elon Musk@elonmusk
We are improving the 0.5T Grok foundation model V8 (public version 4.3) every few days. The 1.5T V9 just finished training (incorrectly called pre-training) and is a major upgrade. Next, we are adding the Cursor data in supplemental training (others call this mid-training), then SFT and RL. About 3 or 4 weeks to release. This will be a banger.
English