Du

@ooobsess 看你手机。定位也要改英国。因为这张卡本身就是 esim 卡。

@Dugdtv 好的，可是我的是安卓不支持esim吧

一天天的真吓人😨 我的giffgaff 卡一直安在手机中，就是一直开着，平时经常接码还发过好几个短信，用了半年了吧，怎么这次没有收到暂停邮件？？？ 这个手机号绑定了超多应用，我都记不住多少了🥺

@imwsl90 为啥不买个 海外 iPhone 呢？

好了，我计划买一台红米k90 pro max手机作为备用机 有准备开始用雷子的手机了，快打醒我 我还没下单！！

@NoraX2026 跨境 11k 在深圳确实不高。但是如果要新人的话，适合去锻炼下。

朋友招聘哈哈！ 感兴趣欢迎找我 超nice！

@ooobsess Esim 可能会好点。并且长期全局+无中国软件，干净环境。平时多用用英国家宽节点。

@Dugdtv 不是，实体卡

@MrGafish 我一直都在用插手机上再用。不过我一直都开着代理。目前还没有收到邮件，希望能躲过一劫。我几乎都用这个卡开的。

我有2张 giffgaff 卡，一张全天待机，一张全天关机 目前都正常使用，没收到邮件提醒 继续观望下这次的风波

@jaylenngx 我在闲鱼买的。700 块。

准备搞一台128G存储的Se三代做主用机了！

@Vincent_AINotes 说的很对，我非常赞同。只是 1.8 的 14 寸没有了，不然就是理想中最划算的。满血加的价对于提升来讲不是特别大。

2026 年买 Mac 跑本地 AI，别只看芯片型号，先看统一内存，再看内存带宽。 结合目前这批本地模型，我的结论很明确： 对大多数人来说，M5 Pro 48GB 是最合适的一档。 跑 Qwen3.6-27B 的 Q6 量化，模型装进去以后，还能给系统、上下文和缓存留出空间。价格又比 M5 Max 低一档，不需要为了本地 AI 直接冲到最高配。 几个容易被误导的地方： ① Qwen3.6-35B-A3B 不是“只占 3B” 它是 MoE 模型，推理时大约只激活 3B 参数，所以计算压力比较低。 但完整的 35B 权重依然要装进内存。 24GB Mac 可以尝试 Q4，但基本属于刚好塞下，后台程序一多、上下文一长，就可能开始吃交换空间。想用得舒服，32GB 起步，48GB 更稳。 ② 8GB Mac 不是完全不能跑模型，但不用考虑这批 27B 以上模型 小参数、低量化模型还能运行，但能力和上下文都会受到明显限制。 买 Mac 专门跑本地 AI，16GB 只能体验，24GB 开始能用，48GB 才比较从容。 ③ M5 Max 不是没用，但 48GB、64GB 版本容易卡在中间 它的内存带宽比 M5 Pro 高很多，模型生成速度也会更快。 问题是价格上涨后，内存容量却没有拉开足够差距。既没有 M5 Pro 48GB 划算，也碰不到 M3 Ultra 256GB 能运行的超大模型。 真正有价值的是 M5 Max 128GB： 既要移动办公，又要跑 70B 级模型、大上下文或者多个模型，才值得考虑。 ④ 想本地跑 DeepSeek-V4-Flash，直接看 M3 Ultra 256GB 这类模型总参数接近 300B，即使量化后，对内存的要求依然非常高。 96GB 很难装下合适的量化版本，128GB 也只能尝试更激进的低比特量化。真正想稳定运行，M3 Ultra 256GB 才是比较现实的选择。 所以，2026 年买 Mac 跑本地 AI，我会这样选： 大多数人：M5 Pro 48GB 必须兼顾便携和大内存：M5 Max 128GB 专门在桌面端跑超大模型：M3 Ultra 256GB 最容易买亏的，反而是 M5 Max 的 48GB、64GB 版本。 买之前先确定自己到底要跑多大的模型。 内存容量决定能不能跑，内存带宽决定跑得快不快。

@elshayib_ @iuditg What package are you subscribed to?

@iuditg I strongly suggest grok also 4.6 and 4.7 are coming in a couple of weeks and the limits are great I really never hit the weekly limit

I love Codex but I am thinking to move away to Kimi K3 or Grok 4.5 due to limit issues. Codex is incredible to work with but they drain out so quickly and baby sitting model between Luna, Terra and Sol is painful. Looks like the reset era is over, and Codex team will do it at their own convenience. My Codex barely lasts for 2 days on Sol 5.6 medium. How has been your experience?

I really don't get it - why is my GPT Pro using up so much data so quickly? I'm a $200 member! What's going on? And my workload has dropped massively now.

@roboticjoey Aquarius actually has $1000? That's awesome!

Anyone that likes this post will receive their share! Reply with your Zodiac Sign: Aries: $3,000 Taurus: $3,000 Gemini: $1,200 Cancer: $7,000 Leo: $5,000 Virgo: $6,000 Libra: $1,000 Scorpio: $2,000 Capricorn: $800 Aquarius: $1,000 Pisces: $900 Sagittarius: $4,000

@Keywest_Felix I think I need $600 to subscribe to Grok for two months. Thanks.

How much money do you need? Maximum : $15,000 I will pick 3 followers

@zhongxingyuyes @luoling8192 这是 codex++么？

@luoling8192 一车内置的面包人，或者一个又主agent自己捏造的面包人

我想要有一个大的 agent，能够帮我派生出「不同模型」的 sub agent，让他们各自去调研或者是 review 我的代码，最后能够整合在一起。有没有这样的软件呢？

@jturntdev I feel like resetting is just a process of taming users now. In 5.4, you couldn't even use it all up, and in 5.5, you still needed tons of time and concurrency just to barely use it up. But everything changed right before 5.6 was about to release.

Sorry but Codex Usage limits are terrible now? Yesterday, i only used 5.6 sol high, i let it running in a /goal, configured cheaper terra/ lunar subagents. I come back 40% of my total weekly usage is gone? ( 20x) BTW i was using multiple 5.6 Sol xhigh + Max, all day everyday, last week. Id rather get no resets, steady and fair usage, as opposed to getting resets, but every time they do, they reduce our quota. They also removed the invite friends for banked resets. Anyone else noticing this? Just gets worse by the day.

@wdn105430 @Vinson1101 你试试用 5.3 就知道了，他除了快，一无是处。

@Dugdtv @Vinson1101 为什么，不懂！ ？

最近codex没有额度的同学，可以尝试一下在命令行终端输入，别浪费了这个额度： codex -m gpt-5.3-codex-spark -c model_reasoning_effort="medium"

@easywaydata 首先应该抑制自己炫耀的心情和思想。其次，为什么要试图教会一个傻子做事？

看着同事手搓Excel, 真想告诉他可以用codex 又怕告诉他后 他问我怎么翻墙，怎么注册，怎么安装，怎么... 我不想说了...

@patloeber I don't agree with them. I think the Ultra membership is definitely worth the money. Not all models are for programming services. But if you're going to be user-oriented, you should focus on one thing and give users exactly what they want.

feedback wanted! Gemini 3.6 Flash & 3.5 Flash-Lite: our goal was to push those models a lot on real world tasks & efficiency, what's your honest feedback after 24h?

@thsottiaux I really want to, but what can I do?🤔

From Science Fiction to Science Reality. Join the team if you want to work on some of the coolest and most impactful technology.

@pashmerepat If you're that scared, I guess I could reluctantly help you use it. How about you give it to me instead? Thanks.

i have a fear that i’m not using the infinite compute openai gives me as a MTS to its fullest potential i’m doing about 20B tokens a day what would you do with an unlimited token budget?