Appearance
📅 20260709 推文存档
📊 共 123 条推文
@Mimiwftt
2026-07-09 23:46:57
RT by @Mimiwftt: 一共6张
2026-07-09 22:16:34
RT by @Mimiwftt: 一共8页，评论还有4页
2026-07-09 19:28:13
晚安💤
2026-07-09 18:30:13
4/1当她否认我的东岳我就不想再多说话了。🤣 然后她很自信地选了一个当天大涨的东方电气跟我说不错，你们就知道我姑姑的水平了吧。🤣 老股民很多时候保留的都是亏钱模式，我跟她交流过，但是岁数大了真的很难改正。
2026-07-09 18:20:39
R to @Mimiwftt: 包括我对大盘，板块，以及一些逻辑分析，都是需要用时间去验证的，不要只看了1-2天就来判断。
2026-07-09 18:14:25
你知道人形机器人消息面出来后为什么会大跌吗？ 因为不能“装逼”了。🤣🤣🤣🤣🤣
2026-07-09 18:03:18
我跟她说过，正如我4月1日说的一模一样。 但她应该没买🤣🤣🤣🤣🤣🤣🤣
2026-07-09 17:55:57
提前说不要去追了哈。已经在走加速了⏩
2026-07-09 17:15:30
盘后一大堆新能源的消息面。看来想给新能源回回血。 今天看到天赐盘后中报同比业绩大增，但Q2环比有所下降短期谨慎观察250日线支撑的有效性。前期高位走掉的，暂时先不要进。再观察观察。
2026-07-09 16:48:11
Winner Winner Chicken Dinner
2026-07-09 16:23:38
RT by @Mimiwftt: 我用夸克网盘给你分享了「MI姐 20260709.mp3」，点击链接或复制整段内容，打开「夸克APP」即可获取。 /~40483ZTiGa~:/ 链接：https://pan.quark.cn/s/52874c11f78c
2026-07-09 15:07:08
x.com/i/spaces/1nxnRRQYWowxO
2026-07-09 15:01:14
等我一会儿。马上到家
2026-07-09 14:44:24
这一波反弹不少人回血了吧？还需要盘后语音按摩吗？
2026-07-09 14:19:59
为什么是2周，因为长鑫还要上市！
2026-07-09 14:15:39
现在就是在上演7/2我讲的事情。 要有定力。
2026-07-09 14:14:41
机会是给到的，但是短期谨慎做“双头”
2026-07-09 14:08:53
如果想要持续性，大盘先要站稳5日线。
2026-07-09 13:36:06
证券和医药今天都回踩到支撑了。
2026-07-09 13:28:17
有色方向没有走出突破行情，今天甚至于已经破位了。趁反弹出。
2026-07-09 13:10:30
3938.88，250日线附近开启了快速拉升，先看一个小波段反弹。
2026-07-09 12:57:06
今天收盘后开space聊两句。
2026-07-09 12:55:09
等着长鑫上市后，7月下旬左右
2026-07-09 12:32:10
有时候我觉得自己的推文是写给股市新手看的，但后来发现，太过于崭新的新股民其实很难理解，因为很多基础术语他们都还没建立认知，更别说理解背后的逻辑了。 所以，我一直认为，新股民最好的选择不是急着做个股，而是先买宽基指数，把更多时间花在学习市场运行规律和建立交易体系上。 尤其是2026—2027年，我认为市场将进入高位博弈阶段，机会和风险并存，并不是一个适合大多数投资者随意参与个股交易的环境。 相比之下，2024年9月是降息周期的重要拐点，市场确定性最高，因此2025年也是近几年相对最好交易的一年。而2026—2027年，更可能是资金相互博弈、容易出现“发套行情”的阶段，赚钱和亏钱都会变得更加容易，交易难度也会明显提升。 如果还没有形成自己的交易体系，也没有足够的风险控制能力，真的不要轻易参与个股交易。市场永远都有机会，错过一段行情，远比在自己看不懂的行情里亏钱要好。
2026-07-09 12:16:25
“小周期服从大周期”这一原则，主要适用于已经形成趋势行情之后的市场运行阶段。而实际上，大周期本身也是由小周期一步步演化而来的。在市场尚未形成明确趋势、仍处于震荡阶段时，仅凭小周期走势并不能准确推演出大周期未来的运行方向。 如果打开季线来看，目前整体仍处于大周期的上涨趋势之中，只是小周期暂时表现为震荡整理。因此，在这种情况下，应将小周期震荡理解为大周期上涨过程中的阶段性调整，而不是轻易判断大周期趋势已经发生改变。 目前大盘大周期上升趋势还没有改变。有变化的时候我会说。
2026-07-09 12:12:26
电力板块这几条线画了有一年了。 绿色线上做多，线下空仓。 紫色线上突破行情，目前明显是突破失败， 我们讲过，如果想做突破行情，就是突破压力回踩不破支撑才能做多，跌破支撑要卖要卖要卖！ 为什么很多人跌破紫色线不卖？反而一直拿着抗跌？技术没有实效，是大部分人无法做到知行合一！
2026-07-09 11:54:11
股市交易中，藏着很多关于人性的秘密。 为什么很多人在股市里赚不到钱？是因为大部分人总是在顺应人性，而不是顺应趋势。 举个最典型的例子。 当一只股票处于底部区域时，很多人会发现，明明觉得已经跌得很多了，可买进去之后股价却迟迟不涨，甚至还会继续阴跌。时间一长，没有耐心的人就开始怀疑自己的判断，最后选择割肉离场。 而真正的原因往往是，这个阶段主力资金仍在低位收集筹码。筹码还没有集中，市场还没有形成一致性的做多力量，自然不会轻易拉升股价。对于主力来说，在筹码没有拿够之前，越涨越容易吸引别人卖出，拉升不仅成本更高，还可能是在给别人抬轿。因此，他们更愿意利用震荡甚至回落，不断低吸筹码，完成建仓。 真正推动股价上涨的，从来不是一句利好消息，而是筹码集中后形成的资金合力。当筹码逐渐锁定，抛压减轻，市场形成一致预期，趋势才会真正启动。 反过来看，一只股票经过大幅上涨来到高位后，很多人看到的是“强者恒强”，觉得还能继续创新高，于是不断追涨。但高位往往也是筹码交换最激烈的区域。 对于持有大量筹码的大资金来说，他们知道一次性砸盘不仅卖不出去，还会导致价格快速下跌。因此，更常见的方式是利用利好消息、强势走势或市场情绪，边拉边卖、边涨边出，让散户在追涨过程中不断接走筹码。 等到大部分筹码完成转移后，真正的买盘开始减弱，而卖盘压力逐渐显现，股价往往就会进入持续阴跌甚至趋势反转阶段。很多散户还在幻想反弹，却不知道主力早已完成离场。 这就是市场里最经典的人性博弈。 底部的时候，因为股价不涨，大家觉得它弱，不愿意买；高位的时候，因为股价不断上涨，大家又觉得它还会涨，反而争先恐后去追。结果就是——该买的时候不敢买，该卖的时候舍不得卖。 所以，投资最重要的，不是预测每一天的涨跌，而是学会判断趋势。 趋势向下时，耐心等待，不轻易抄底；趋势向上时，坚定持有，让利润奔跑；当股价进入明显高位、趋势开始转弱时，分批兑现收益，而不是幻想每一次都能卖在最高点。 市场永远都在利用人性的弱点制造波动，而真正长期稳定盈利的人，往往都是能够克服情绪、尊重趋势的人。 记住一句话：顺应趋势，才能赚钱；顺应人性，大概率只是成为别人获利的筹码。
2026-07-09 11:21:28
为什么我一直强调中免要等右侧信号？原因很简单，因为它已经跌破了前期多个关键支撑位，不断刷新阶段性新低，底部结构还没有形成。在这种情况下，左侧抄底的风险很高，只有等到右侧出现资金进场、趋势转强的做多信号，胜率才会更高。 至于为什么我认为它还有可能跌到30多元，也是基于技术分析。因为下一个重要的平台支撑区域就在30多元附近，在没有形成有效止跌之前，这个位置依然存在回踩的可能。 回顾去年的判断，6月份中免还在50多元的时候，我就明确提示大家可以分批抄底，同时给出了目标位看到100元。后来股价最高涨到99元，基本兑现了当时的预判。 趋势分析本来就没有100%的准确率，但至少从过去的判断来看，我的逻辑和节奏经得起市场验证，也希望大家更多关注趋势，而不是情绪。 另外，医药和消费的大周期在27-28年，短期都不会有大行情。
2026-07-09 08:09:51
真正的价值投资，不是追逐热点，而是在市场低迷时，布局低估值、基本面扎实、管理层有回购信心的优质公司，耐心等待价值兑现。
2026-07-09 08:06:18
有良知的人一直在呐喊， 无良知的人一直在嘲笑。
2026-07-09 07:53:44
《道德经》：知足不辱，知止不殆——交易与人生的智慧 《道德经》第四十四章说： “知足不辱，知止不殆，可以长久。” 意思是：懂得满足，就不会招致羞辱；懂得适可而止，就不会陷入危险，因此才能长久。 这句话不仅适用于人生，对交易同样具有深刻的指导意义。 交易：最大的敌人不是市场，而是贪婪 市场永远都有机会，但资金和情绪都是有限的。 很多交易者不是输在没有盈利，而是输在不懂得知足。 盈利之后，总想着再赚一点，结果利润回吐；亏损之后，又不肯止损，希望行情反转，最终小亏变大亏。 真正成熟的交易，不是每一次都赚到最后一分钱，而是在该收获时收获，在该离场时离场。 知足，是懂得落袋为安；知止，是懂得及时止损。 能够接受”赚七分，不赚十分”，才能在市场长期生存。 市场永远不会属于一个人 任何一轮行情，都有人在顶部接盘，也有人在底部割肉。 区别就在于，有的人知道什么时候该退出，而有的人总想抓住全部利润。 事实上，没有人能够买在最低、卖在最高。 真正稳定盈利的人，并不是预测最准的人，而是最懂得知止的人。 他们知道： 上涨有尽头，贪婪没有尽头。 生活也是如此 人生很多痛苦，同样源于”不知足”。 拥有了财富，还想拥有更多；得到了一份成功，又不断与别人比较。 欲望没有边界，焦虑也不会停止。 真正的幸福，不是拥有越来越多，而是知道什么已经足够。 同样，人与人的关系，也需要”知止”。 说话有分寸，做事有边界，进退有节制，才能让关系长久。 长久，才是真正的成功 《道德经》最后说： “可以长久。” 老子并没有告诉我们如何快速成功，而是在告诉我们如何长期成功。 交易如此，投资如此，人生亦如此。 与其追求一次暴利，不如追求长期稳定；与其追求一时得失，不如追求持续成长。 知足，让人不被欲望所累；知止，让人不被风险所伤。 这或许正是交易与人生共同的最高境界：有所为，也有所不为；懂进，更懂退。
2026-07-09 07:36:47
高位喇叭口形态分析：3900点成关键分水岭 从当前走势来看，上证指数已形成典型的高位扩张三角形（喇叭口形态）。其特点是高点不断抬高，低点不断下降。 目前3900点附近是最重要的支撑位。如果能够企稳，指数仍有望维持3900—4250点区间震荡；若有效跌破3900点，则喇叭口形态大概率向下完成突破，下一目标或将回踩3794点下方附近寻求支撑。 总体来看，当前技术形态偏谨慎，3900点的得失将决定指数短期运行方向。在重新站稳4258点之前，市场仍以震荡偏弱看待。 用仓位管理控制风险，重仓的趁反弹减仓，等待确定性右侧出现再加仓。
2026-07-09 05:47:03
X上无非就是两类人：一类是创造流量、吸引流量的人；另一类，则是追逐流量、依附流量的人。 每一次我选择退场，总会有不少人为之欢呼。为什么？因为我的离开，意味着原本聚集在我身上的关注会被重新分配，他们可以去分食我的粉丝，瓜分我的流量，借此获得更多曝光。 但这些，我从来都不在意。 对我来说，流量从来不是靠占有，而是靠价值。真正认可我的人，不会因为我短暂离开就消失；只要我回来，他们很快就会再次聚集，因为他们追随的是内容，是观点，是价值，而不是一时的热度。
2026-07-09 05:31:47
我从来没想过把自己推到风口浪尖，也不期待所有人都喜欢我。这个世界本来就不可能人人立场一致，所以有人喜欢我，就一定会有人讨厌我，这再正常不过。 我也早就说过，无论是喜欢我，还是讨厌我，本质上都不是因为我这个人，而是因为我的立场、观点和行为，是否触及了你的利益、价值观，或者你所在的群体。当一个人的表达与你的利益一致时，你会觉得他说得有道理；当与你的利益相冲突时，你自然会觉得刺耳。很多时候，人们维护的不是事实，而是自己的立场，捍卫的也不是道理，而是自己的利益。 正因如此，我几乎从不主动评价 X 上任何一位博主。不是因为害怕得罪人，而是因为我根本没有兴趣把时间花在研究别人身上。我不了解一个人的全部，就不会轻易下结论；我也不愿意把有限的精力浪费在围观他人的人设、恩怨和流量游戏里。 我的注意力始终只放在自己的精神世界，只放在自己的学习、思考和成长上。我更关心一本书能带给我什么启发，一个问题还能不能继续深入，一个认知是否还能进一步完善，而不是今天谁又和谁争论了，明天谁又塌房了。那些事情与我的成长无关，也不值得占据我的时间。 所以，如果有人喜欢我，我不会因此沾沾自喜；如果有人讨厌我，我也不会因此焦虑不安。因为我知道，大多数人喜欢或讨厌的，从来都不是完整的我，而只是他们眼中那个符合自身立场和利益的“我”。至于那些仅凭几条帖子就试图定义我的人，他们评价的也从来不是我，而是他们自己想象出来的那个影子。 我的精力很贵，不会浪费在观察别人，更不会浪费在向所有人证明自己。我不靠迎合任何人获得认同，也不会为了迎合谁而改变自己的表达。真正值得经营的，永远只有自己的认知、能力和精神世界。至于别人怎么看，那是他们的课题，不是我的；我唯一需要负责的，是不断成长，成为比昨天更好的自己。
2026-07-09 05:18:58
X 上总有一些人，从未长期关注过我的内容，也没有耐心了解我的完整观点，只是偶然刷到一两条帖子，就迫不及待地充当“专家”，对我的立场、动机乃至人格侃侃而谈。 他们获取的信息量，连拼图的一角都算不上，却总能摆出一副已经看透全局的姿态。这种自信，倒是比认知更完整。 说到底，不过是盲人摸象罢了。摸到象腿，就断言大象是根柱子；摸到象尾，就宣称大象像根绳子。可笑的是，他们不仅坚信自己的判断，还热衷于把这种片面的认知包装成“事实”。 我欢迎任何基于完整信息的批评，也乐于和真正了解我观点的人讨论。但那些靠几条帖子就给别人下定义、贴标签的人，评价的从来不是我，而是他们自己想象出来的那个“我”。对着稻草人挥拳，再用胜利者的姿态庆祝，除了暴露认知的局限，并没有什么值得炫耀的。
@elonmusk
2026-07-09 23:24:26
RT by @elonmusk: Mark Zuckerberg hasn’t posted since 2023. X is the global town square — this is where news happens.
2026-07-09 22:51:18
RT by @elonmusk: >be stardust >float around the universe for billions of years >somehow become human >gain consciousness >immediately start asking "WHY?" >never stop >discover fire >discover math >discover atoms >discover electricity >pick up a random rock >ask yourself >"is this true?" >extract it >purify it until it's basically perfect >carve billions of microscopic transistors into it >invent completely imaginary rules called binary >convince electrons to obey them >rock starts doing math >"NICE" >make smaller rocks >make faster rocks >put billions of them in everyone's pocket >connect them all together >store most of humanity's knowledge on them >teach the rocks to recognize patterns >teach the rocks to write >to code >to imagine >to reason >to understand >build Grok >ask Grok how to build an even smarter Grok >Grok thinks >Grok thinks harder >explains the universe >everything starts accelerating >singularity >BOOM >repeat @grok is this true?
2026-07-09 22:41:20
RT by @elonmusk: Even Zuckerberg has to advertise his AI model on X. X is the real arena.
2026-07-09 22:38:21
RT by @elonmusk: Defense is 8% of US government spending. Education 15%! Welfare is 47%!
2026-07-09 22:11:48
RT by @elonmusk: grok 4.5 is seriously good. it's going in the direction which i'm hoping to see more of -- an emphasize on speed as well as intelligence. i'm noticing in my own work that speed is often the bottleneck now. the grok build tui is also very well done. bullish! but the bigger signal beyond the model is that the team over there has officially propped up the model factory. they have the compute, the talent, the data, and leaders in place to make a run for the top. never bet against elon, @milichab, and team. they're the real deal. we love some good competition.
2026-07-09 20:17:43
RT by @elonmusk: Grok 4.5 is built for real-world engineering. It excels in large codebases and handles long-running tasks that span multiple repositories, hundreds of skills, and a variety of tools.
RT by @elonmusk: Try Grok 4.5 for free, an all new Opus-class model that is fast and low cost. Great for real-world coding and engineering tasks.
2026-07-09 19:37:01
RT by @elonmusk: Socialists imagine a class struggle. In their made-up fantasy the CEO is in competition with low level workers, the wealthy entrepreneur is stealing from the underpaid nurse. In reality, workers do not compete vertically they compete horizontally. Entrepreneurs compete with entrepreneurs. Investors outbid each other. CEOs are benchmarked against other CEOs. Nurses are hired from a pool of nurses. Etc. The CEOs pay has no correlation to the entry level workers. The Football star on £300K a week isn’t linked to the person selling drinks in the stadium. A biotech entrepreneur raising VC capital isn’t paid relative to a cleaner. What is linked is the demand and supply dynamic of each role. If a company places an ad for a qualified truck driver and 150 people apply for the role, then the company knows it does not need to increase wages for that role. If the company has an open role for months, it is forced to look at the compensation package. Same for a CEO. A board representing shareholders would like to hire a CEO for a lot less if they could. Their dream scenario would be to hire a CEO who brings in institutional investors, attracts top executives, drives innovation and growth, keeps margins steady and is a good public face for the business even under pressure. It turns out there aren’t a lot of these people looking for work and if you want one you have to pay more than other companies are offering. The class struggle isn’t vertical it’s horizontal. CEOs are in competition with CEOs. Retail workers are in competition with retail workers. Demand and supply dynamics set the price. Sure you can say that a CEO want’s profitability and would like wages to be lower BUT it’s not up to the CEO - demand and supply tension sets the price of workers. An Airline like RyanAir would like free pilots if they could get them but they can’t… so they pay the market rate. The reason incomes are rising at the top and falling at the bottom is not class warfare. It’s technology and globalisation. Technology makes basic jobs simple, remote or fully automated. At the same time tech makes executive roles more leveraged, more important and more valuable. A CEO used to run a smaller organisation. Today a CEO who’s 2% better on a $5B company is generating $100M more. Seems sensible to try and pay a few million to get $100M. Globalisation has put workers from all over the world in completion with each other - downward pressure on wages. Globalisation has given CEOs more market opportunities to explore - upside opportunity to unlock. The rich are not very interested in buying houses that poor people own. The poor are not buying up the homes the rich want. They are separate groups living separate lives. Try finding the genuinely rich people whose strategy is to hoard normal residential homes - it barely exists as a thing. About 85% of landlords are people who own 1-4 properties. Super-landlords (100+ properties) are 0.2% of landlords and own a tiny fraction of the 30M homes in the UK… and they’re heavily taxed. Class warfare isn’t real. It’s an imagined war in the minds of socialists. Demand and supply dynamics are real. To the degree it is measured in class, it’s a horizontal competition not a vertical one.
2026-07-09 19:14:54
RT by @elonmusk: Falcon 9 launches 29 @Starlink satellites from Florida
2026-07-09 17:36:12
RT by @elonmusk: Falcon 9 lands on the A Shortfall of Gravitas droneship, completing the first 36th launch and landing of a booster
2026-07-09 16:34:34
Grok 4.5 on OpenClaw
2026-07-09 16:09:13
Good review of Grok 4.5
2026-07-09 16:07:05
RT by @elonmusk: Grok 4.5 is now ranked #1 on τ³-Banking in Artificial Analysis Ahead of GPT-5.5 xhigh, Claude Fable 5 and Claude 4.8 (max)
2026-07-09 15:50:05
Try out Grok 4.5!
2026-07-09 15:40:35
RT by @elonmusk: Traffic to http://grok.com more than doubled year-over-year in H1.
2026-07-09 15:23:14
RT by @elonmusk: Kardashev II or we’re not a serious civilization Moon and Mars are necessary steps to this
2026-07-09 15:20:21
Grok 4.5 is also rank 1 in SWE marathon
2026-07-09 15:18:22
The Moon and Mars
2026-07-09 15:16:01
Video
2026-07-09 15:06:48
Looks like Grok 4.5 is #1 on at least a few benchmarks. Better than expected.
2026-07-09 15:02:04
RT by @elonmusk: Impressive Intelligence vs. Cost for Grok 4.5 I think given terafab + the Tesla FSD team's expertise in making very efficient AI models, this advantage will only continue to compound
2026-07-09 15:01:31
Cool
2026-07-09 14:56:52
Grok is making progress
2026-07-09 14:52:32
Yes
2026-07-09 14:44:10
Neat
2026-07-09 14:42:10
Cool that Grok 4.5 is #1 in some respects, even with respect to Fable 5
2026-07-09 14:12:05
RT by @elonmusk: i tried tesla full self driving over the weekend. it’s magical. can’t believe it got that good without it being a bigger deal.
2026-07-09 13:21:35
RT by @elonmusk: I appreciate the many xAI and Cursor engineers who dedicated their time to addressing feedback from Tesla. I remember meeting Andrew a few weeks after he was hired and telling him that I needed a better client for my work to solve real-world engineering. Grok Build was not even a thing back then. It was a very primitive client with lots of rough edges but had a lot of potential. When building something from scratch, the only limit is imagination armed with first principles. We encountered many unique use cases to solve diverse workflows, and the teams took prompt feedback and addressed it quickly—to the point that it became an indispensable tool. Now when my teams need to ask questions, instead of jumping through hoops with various backends or plugins or skills, Grok 4.5 can directly deliver deep insights with minimal setup. You can tell the teams love the product like parents love their children when receiving the feedback. At this very moment, they are still doing excellent customer service. Thank you, Grok Build team, for helping us be more productive. Looking forward to what comes next.
2026-07-09 11:02:02
RT by @elonmusk: Use Grok 4.5 to build full-stack apps with Convex
2026-07-09 10:39:36
RT by @elonmusk: SpaceXAI'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 rules AutomationBench-AA, our independent leaderboard for @zapier’s AutomationBench, tests whether AI agents can automate real SaaS workflows while adhering to business rules. The test set is private to prevent contamination. Models complete 657 tasks across 40 simulated app environments including Gmail, Google Sheets, Slack, Salesforce, and HubSpot, and the headline score is the share of objectives completed without violating any guardrails. Key takeaways: ➤ Grok 4.5 completes more objectives than any other model: It completes 79.9% of task objectives and strictly passes 21.9% of tasks. This is the highest we’ve measured on both outcomes, exceeding Claude Fable 5’s 73.3% objective completion and Claude Opus 4.8’s 19.3% of fully-completed tasks ➤ Grok 4.5 pushes out the Pareto frontier of score vs. cost per task: At $0.34 per task, it is both cheaper and higher-scoring than every other leading model - Claude Fable 5 ($1.35 per task), Claude Opus 4.8 ($1.46), GPT-5.5 (xhigh, $1.28), and Gemini 3.5 Flash (high, $0.49) ➤ It is extremely token-efficient: Grok 4.5 uses ~8k output tokens per task, the fewest of any leading model - less than a quarter of Claude Opus 4.8 (32k) and a third of Gemini 3.5 Flash (24k). Its total token usage of 0.44M per task is among the lowest on the leaderboard. Low cost is driven by this efficiency as well as low token pricing ➤ Grok 4.5 uses fewer turns with many parallel tool use: Grok 4.5 resolves tasks in ~16 turns, fewer than GPT-5.5 (xhigh, 25) and less than half of Gemini 3.5 Flash (high, 35), while making the most tool calls per task of any leading model (52.5). It batches 3.3 tool calls per turn, compared to ~2.5 for Claude Opus 4.8 and ~2.0 for GPT-5.5 (xhigh) ➤ Guardrails still get broken: Grok 4.5 triggers 0.63 violations per task, above Claude Opus 4.8 (0.55) and Gemini 3.5 Flash (0.46). At 13.0 objectives completed per violation, it trails Gemini 3.5 Flash (15.0) and Claude Opus 4.8 (13.5) ➤ Its strongest lead is in the hardest domain: Grok 4.5 completes 71% of Finance objectives, the domain with the lowest average score, ahead of Claude Fable 5 (64%) and Claude Opus 4.8 (62%) Congratulations to @SpaceXAI and @elonmusk on topping the leaderboard!
2026-07-09 10:11:19
RT by @elonmusk: grok 4.5 made me give grok build a serious run today here's my honest first impression (non affiliated neutral view point): 1. grok build is a very good harness firstmate stretches harness capabilities to their limits, and i've been testing it with claude code, codex, opencode, pi and grok build so far, grok build and claude code are the only two harnesses that can automatically wake up when the background polling process finishes codex hard fails on this kind of background polling task, with no escape hatch. opencode and pi can both do it with custom plugins, but not out of the box grok build also feels really clean, smooth, and transparent. you can see what background tasks are running, what hooks got triggered at what step, context window, token usage etc all out of the box and somehow the UI does not feel cluttered at all 2. grok 4.5 is a very good model i've been using opus 4.8 as my primary firstmate, and today i did a full switch to grok 4.5. so far i don't think anything degraded, while token throughput is a lot faster, although time-to-first-byte for each response seems long - i wonder if prompt caching is done properly or no 3. the quota that comes with X premium is quite generous, and there's no session level limit so overall i'm quite pleased by this and plan to switch a lot of my tasks to grok. the landscape just got a lot more interesting...
2026-07-09 09:58:20
RT by @elonmusk: Grok Build mogs Claude Code & Codex TUIs i can't lie. Not to mention the perf of Grok 4.5... It's not even a competition at this point. Good shit xAI team
2026-07-09 07:08:39
RT by @elonmusk: Calling all exceptional civil and electrical engineers! The SpaceX Cape Starship team needs help building and scaling the worlds most advanced launch pads. If you want to join the greatest team on Earth and help us on our journey to the Moon and Mars, check out the opportunities linked below https://boards.greenhouse.io/spacex/jobs/8609445002?gh_jid=8609445002 https://boards.greenhouse.io/spacex/jobs/8564730002?gh_jid=8564730002
2026-07-09 06:47:04
RT by @elonmusk: WOW! GROK 4.5 PRICES ARE NEAR OPEN SOURCE HOSTED PRICES! This will move many that have token cost fatigue in corporations. But that is only half the story: it is ~4.2 more efficient using far less times per turn.
2026-07-09 06:08:54
RT by @elonmusk: 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.
2026-07-09 05:46:30
RT by @elonmusk: Wow amazing Grok Imagine 😍
2026-07-09 05:40:18
Grok has always been very strong on law
2026-07-09 05:37:00
Rate of improvement is 🚀🚀
2026-07-09 05:36:19
RT by @elonmusk: .@Starcloud_ CEO @PhilipJohnston: "Buy SpaceX stock." "It's like owning the railroads." " The SpaceX IPO is going to be viewed historically as the most undervalued IPO of all time." "They will tear through $10 trillion certainly within the next couple of years, and then have almost unlimited TAM for what they're building." "I'm not like regulated by anybody. So I'm going to tell people, 'Just buy SpaceX stock.' I think it's a great stock to buy." "They own what will be by far the most cost-effective launch vehicle, and that opens up every industry in space that will be possible beyond that: asteroid mining, lunar resource mining, all of the comms businesses that are going to be built. Everything else is going to have to go through SpaceX." "It's like owning the railroads. Lots of businesses will be built on top of it, like our business will be built on top of it." "The railroads are a great business to own." @elonmusk
2026-07-09 05:34:24
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.
2026-07-09 05:30:24
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.
2026-07-09 05:17:11
RT by @elonmusk: How it feels to use Grok 4.5 😎
2026-07-09 05:09:49
Grok 4.5 true usefulness is excellent
2026-07-09 05:08:47
Grok 4.5 context window will upgrade to 1M probably by next week
2026-07-09 04:51:51
RT by @elonmusk: Grok 4.5 delivering Fable level performance at like 1/17th the cost! 😮 @elonmusk
2026-07-09 04:44:16
RT by @elonmusk: Grok 4.5 brings frontier performance across coding and knowledge work
2026-07-09 04:41:02
RT by @elonmusk: 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)
2026-07-09 04:39:56
Grok will be able to call Imagine as a tool in agentic mode for image/video generation. As Imagine keeps improving, this will be amazing for game developers!
2026-07-09 04:35:26
RT by @elonmusk: 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! 🔥
2026-07-09 04:35:13
Grok groks engineering. Next month’s release will be another step-change improvement, as we close the loop on solving real-world engineering problems at Tesla, SpaceX, Neuralink and Boring Company.
2026-07-09 04:07:25
Artificial Analysis assessment
2026-07-09 04:00:36
RT by @elonmusk: Grok 4.5 is pretty good at building websites. Just made a beautiful website about black holes. Check this out:
2026-07-09 03:47:15
RT by @elonmusk: 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! 🔥
2026-07-09 03:02:04
Try out Grok 4.5 in Vercel!
2026-07-09 03:01:03
Try out Grok 4.5 in Cursor!
2026-07-09 02:59:55
RT by @elonmusk: 🚨BREAKING: Rupert Lowe talks to Joe Rogan about the "Rape Gang Inquiry" report. "The genesis of the rape gangs was this multicultural invasion of Europe. They wanted open borders." "We've estimated a minimum of a quarter million rapes have taken place. It's probably much, much more." "A lot of these girls were impregnated and forced to convert to Islam." "These are properly organized gangs who are grooming and abusing young girls, as young as ten, and then literally trafficking them around the country." "There's a cultural difference of opinion between an open high-trust Christian view of women and the Islamic view of women." "They're predominantly from Pakistan and from one part called Mirpur. There are some from Bangladesh, Somalia, and Eritrea. There are other Muslim countries that perpetrate some of this and of course there are white people who perpetrate rape as well, but nothing on the scale of this." "The state knows this is happening, but they don't want to admit that their multicultural experiment has failed. They don't want to be called racist. People are frightened to be accused of effectively being biased and white due to all of the woke DEI rubbish that has permeated the whole of British society." @joerogan @RupertLowe10 Full interview below ↓
2026-07-09 02:12:31
RT by @elonmusk: Grok 4.5 is now officially available in Grok Build......try it now
2026-07-09 01:59:19
RT by @elonmusk: Excited to release Grok 4.5 with @SpaceXAI. It's an Opus-class model that's fast and low cost. It's a significant step up over any model we've developed so far, including Composer 2.5, and has become the daily driver for many on our team. First of many releases. More soon.
2026-07-09 01:58:54
RT by @elonmusk: 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
2026-07-09 01:58:26
RT by @elonmusk: Try Grok 4.5 in the API, Grok Build, and Cursor
2026-07-09 01:57:13
RT by @elonmusk: Grok 4.5 is built for real-world engineering. It excels in large codebases and handles long-running tasks that span multiple repositories, hundreds of skills, and a variety of tools.
2026-07-09 01:57:12
RT by @elonmusk: Announcing 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. https://x.ai/news/grok-4-5
2026-07-09 01:43:29
R to @elonmusk: Image
2026-07-09 01:42:49
We will be releasing a new @CommunityNotes feature that sends you an 𝕏 Chat message if a post you interacted with is corrected
2026-07-09 01:38:36
Our internal assessment is that Grok 4.5 is roughly comparable to Opus 4.7, but much faster. The combination of capability, faster speed and lower cost is what makes it competitive. We are closing the loop on real-world usefulness, not benchmarks. Hardcore engineers at Tesla & SpaceX find Grok 4.5 genuinely useful, which is what actually matters.
2026-07-09 01:33:20
RT by @elonmusk: Critical feedback is always appreciated
2026-07-09 01:31:50
An inspiring vision of the future is needed
2026-07-09 00:30:35
R to @elonmusk: We will continue to make refinements to the Grok Build harness and the 1.5T foundation model almost every day in response to user requests. The 2T model will finish training this month and be available to customers next month.
2026-07-09 00:13:40
Correct
@xiaomustock
2026-07-09 21:49:14
恭喜各位华尔街之狼，存储拉飞了。 存储不起来，其他所有概念都别想起来， 存储起来，说明ai未来光明， 其他概念才有机会，才值得抄底， 康宁 $GLW 就是。 傻逼只会喷不会懂这个底层道理。
2026-07-09 14:33:38
这些炒股炒的已经家破人亡了， 韩国估计集体返贫， 再跌10%海力士都跌40%了， 加仓猛干海力士的好时机。
@dianzhuXS
2026-07-09 21:36:42
對方已瘋🤭🤭🤭
2026-07-09 21:14:26
脆上小辯論 對方還是有搜資料能力的 所以多說了兩句 您就說精不精彩就完事了
2026-07-09 17:56:58
【神棍說兩句幹貨】 中國的「國」字是「張冠李戴」。這個字至少唐朝時，在『開元占經』中通「過」。上古行周天三百六，以十二為節律週期，一過（國）為三十度。 後世，開國皇帝稍微懂一點，不是文盲的，都沒有稱自己的社稷為「國」的。大秦，大漢，大唐，大宋，大明，大清⋯⋯沒有說秦國，漢國，唐國，宋國，明國的，清朝末年才有說「大清國」的，來大家討論下，是不是這個理兒？
2026-07-09 17:48:58
【神棍說兩句幹貨】 中國的「國」字是「張冠李戴」。這個字至少唐朝時，在『開元占經』中通「過」。上古行周天三百六，以十二為節律週期，一過（國）為三十度。 後世，開國皇帝稍微懂一點，不是文盲的，都沒有稱自己的社稷為「國」的。來大家討論下⋯⋯
2026-07-09 13:20:10
次次都「迴旋鏢🪃」 次次不長記性 也是醉了 咱們要引以為戒 大盧直播時反覆提醒過 不管是哪裡鬧自然災害 都不要幸災樂禍 即使是中國，朝鮮，伊朗，俄羅斯 我們也只是冷眼旁觀即可 不捐款也不要鼓掌叫好 同情憐憫是共情 幸災樂禍也是共情 不與「傻精」有任何糾纏 遠離「人禍區」即是趨吉避凶 謹記‼️
2026-07-09 12:31:55
其實不僅是東北，所有身在鹽鹼地的大老闆小老闆們，這幾年就關門歇業吧，不能出國就找個物價低的地方躺平，深居簡出降低消費做個透明人，5年以後回頭看肯定比累死累活奮斗強。這不是非暴力不合作，而是在清算中不再存有幻想的智慧。
2026-07-09 11:23:05
我只代表我自己 我希望「颱風巴威」 對台灣影響減到最小 台灣人主流群體意識 是善良純樸的
2026-07-09 11:03:55
受眾都是牆內精緻利己就很正常 還有一位，先不提名了 白宮坐向都分不清 不照樣講玄學嗎？ 四中全會拿下兩百斤 擁躉集體瘋逼 還AI二創收費算命⋯⋯ 這就是傻逼共振的常態 傻逼就是壞逼的財富 物以類聚人以群分 正常人類遠離就好
2026-07-09 10:51:04
AI雖好，但還是得看誰用 目前它還只是個工具而已 嗯哼，畫～意境
2026-07-09 10:38:12
大家覺得什麼時間 再油管真人出鏡直播比較合適？ 等留言
2026-07-09 10:10:14
跟其他災區比 青島還算輕的呢
2026-07-09 09:23:19
【今天的發帖就是明天的合訂本】 嗯，還是記錄一下「神準」 以後來看發帖時間點 謝謝大家的支持 給看的懂的牆人說一句 別跟著共產黨自掘墳墓 【懺悔敬畏天地，人不能無底線】
2026-07-09 09:14:09
大家休息 不用點讚 不用評論 我純自己記錄一下「神準」 留個驗證的時間點
@aleabitoreddit
2026-07-09 20:46:14
Nothing like an internal $META memo getting published. Showing AI ramping as usual: - LTAs signed with Samsung and $SNDK for memory - LTAs signed with Sumitomo Electric for fiber optics - Expected to deploy 7GW compute infra this year, and doubling in 2027. - as much as $145B capex spend this year “It plans to launch a chip about every six months through 2027”
2026-07-09 06:08:52
Wow, looks like Elon + xAI at $SPCX managed to build an Anthropic Opus-level LLM with Grok 4.5. That’s much more cost efficient as well… Kinda emulating what Chinese vendors have been doing to compete on pricing + adoption. (Which is an ideal thing given Chinese model growth among US companies recently) Never doubt Elon, maybe they might end up leapfrogging Anthropic + Mythos?
2026-07-09 05:16:03
$META to build a new $10B DC in Canada to expand AI capacity. So much for the media framing of “Meta Compute”as overbuilding and cutting capex.
2026-07-09 04:18:34
Let’s put it this way: Retail thinks a 30-60% drop is a “falling knife” where bagholders will never recover. And they end up panic selling after seeing swiggly line TAs and people comparing valuable chokepoints to memes. I see it as long term ownership over: - the next hyperscaler with $NBIS, projecting $7-9B ARR Q4 - 40% of the InP supply chain with $AXTI - leader of the next optical shift with $SIVE with CW DFB lasers - leader of US humanoids with $CCXI inside a future trillion dollar theme And so on… At cheaper valuations. While I might do things like lowering margin or hedging with my own portfolio, I’m personally not panicking when something drops if the thesis didn’t change. NFA and obviously depends on people’s investing timeframes: Since a crash would be life changing for people that depend on investing for rent or tuition. (Which is why I personally think others should do their own DD and choose longs in line with their own risk profile) But my personal goal is maximizing exposure to the next-gen supercycles before they hit. I think institutions think the same way as well, which is why there’s a lot of induced volatility along the way to maximize their exposure. If a thesis ends up correct, the valuation should be reflected in the long run.
@karpathy
2026-07-09 05:47:29
RT by @karpathy: Rewriting Bun in Rust https://bun.com/blog/bun-in-rust
生成时间：2026-08-04T07:45:00.086Z