darktalker

@CarolGLMs Where is linux version?

More good news 🚀 while GLM-5.2 is rolling out to GLM Coding Plan users. ZCode 3.0 brings stronger coding capabilities, plus extra quota and free trial offers.

@zcode_ai @gkhnyldrmlr Invite me pls

@gkhnyldrmlr Good news: we do have a Linux version in development. If you’re interested, we’d love to invite you to the beta and hear your feedback.

GLM-5.2 is now fully available for GLM Coding Plan users. ZCode 3.0 is deeply optimized for GLM-5.2, bringing stronger Agent task execution, better long-context coding, and the new Goal feature for managing larger development objectives from planning to completion. Coding Plan subscribers get 150% usage quota inside ZCode. New users get 5 days free with 5M tokens per day. Download: zcode.z.ai

@zcode_ai My too, interested in Linux version

@bruc3van 不支持linux

想要试用 glm-5.2 又买不到/没有 glm-coding-plan 的兄弟，可以试试智谱的 coding 工具 ZCode 。 新用户连续免费使用 5 天 GLM 旗舰系列模型用量 现在越来越多模型厂商下场做 coding 工具，或者说通用办公智能体了（形态上往桌面 GUI 方向发展，覆盖更多非编码人员）。 第三方Agent厂商除了模型选择自由这个优势之外，还有什么竞争力？ zcode.z.ai

@MiniMax_AI Please increase your 5h tokens limit.

Who’s claiming follower #100,000? 🧋👀💃

@cherry_mx_reds Pi + deepseek if vision is not needed, otherwise pi+ minimax m3

Please give me your best combo of harness + Chinese LLM and I will give it an earnest shot. It has been a while since I've experimented with them.

@MiniMax_AI M3 is amazing, but pls fix the cache hit rate issue

Weights on Friday 🫶

@RyanLeeMiniMax Recently cache hit rate is very low, less than 70%, is that normal?

🔥 MiniMax-M3 just got faster. Thanks for all the excitement around MiniMax-M3 — the response has been far beyond our expectations. Last night, we rolled out a major inference upgrade: 🛠️ Fixed an issue that could occasionally produce abnormal tokens 💾 Increased memory and improved cache efficiency 🚀 ~50% higher throughput, with most users now seeing 50–70 TPS You should notice a much smoother experience today. More optimizations are on the way. ❤️

@uzairansar Using minimax M3 for about one week already, very impressed, especially for frontend development where you need quick feedback about designs, audio etc. And it's cheap, I don't mind to do more 1 or 2 rounds comparing to Claude.

MiniMax M3 just got added to DeepSWE. As expected, Kimi k2.6 remains the goat of all Chinese models.

@mranti @riba2534 我也是讯飞，看来他们没跟进

@riba2534 我装了2.1.156，不行，

Claude Code 的 2.1.156 发布了，修复了API端点的不兼容问题，现在所有的国产厂商，端点已经都可以支持使用 dynamic workflows 了，对于几百个大型的流程编排，还是使用便宜的Deepseek之类的玩吧🫡

@CarlZha In Paris, the 14 metro line is driverless too

People were scared to use Chengdu metro line 13 because it's driverless. Metro's solution is to put a guy who sit in front to pretend he is the driver. There's no controls and he is not allowed to play w mobile phone. His job is just be there to make people comfortable 😭

@luoleiorg deepseek出图像识别后，就没其他国内套餐什么事了

阿里云 Coding Plan 上月收紧、推出 Token Plan 后，qwen3.7-max、kimi-k2.6、glm-5.1 这些新模型都只上了 Token Plan，DeepSeek 全系和图像生成 Coding Plan 也没有。之前 Coding Plan 营销页那句「支持最新最强编程模型」也被悄悄去掉了。感觉是在刻意让 Coding Plan 冷死亡，逼老用户逐步放弃续费🙄

@joaocorreia_dev @LottoLabs Ollama cloud has larger quota vs price. If you are a heavy user, go with ollama

@LottoLabs Between opencode go and ollama cloud? Whos better?

Update on Opencode Go It’s great value for $5/month, there’s really no reason not to do the first month. At $10/month it’s still good value and gets you access to all sota OS models. You can’t daily drive it without hitting limits on the big models but w/ Kimi x3 you won’t hit limits unless you’re insane. Overall highly recommend the first month, then make your own decision.

@BruceBlue 一个问题吃我70%的五小时量，太会抢了

火山方舟的 Coding Plan 这么便宜？买一个 Lite 实测一下 Kimi K2.6🤔 智谱 GLM Lite ¥44 才 80 次/5小时，MiniMax Starter 要$10，Kimi 自家最低档也得 $15 起步…… 而火山 Lite 只要¥40（不到$6），每5小时约 1200 次请求！Pro ¥200（不到$30）5倍用量。 我主要用 Kimi K2.6 在 Hermes Agent 里测试，感受如下。 🔹国内优秀模型不用抢：套餐内包含 Kimi-K2.6、GLM-5.1、MiniMax-M2.7 等比较有优秀的模型； 🔹工具全兼容：Cursor、Claude Code、OpenClaw、Hermes Agent等，额度共享； 🔹用量消耗：昨天订阅后，轻度使用（2个Agent系统Code Review + Bug修复）花了10%额度。 如果狠狠用几天，一周不到大概率用完；重度用户建议直接上 Pro； 🔸缺点：偶尔会因为“用的人太多”触发 429 限流（尤其是Kimi-K2.6和GLM-5.1），Auto 模式会智能避开高消耗模型，可能会调用字节的豆包模型🐶。 ✅老实说：Kimi K2.6 在方舟上表现极强，Bug 定位/修复速度和解决问题的能力保底优于 Claude Sonnet 4.6，达到了 Claude Opus的～90%左右。 👉新用户首单有 9 折：volcengine.com/L/a4d2ATBQME4 #kimi #GLM #MiniMax #HermesAgent

@mranti 火山就是骗子，一个问题可以烧我50的五小时额度，相比下来讯飞简直太良心了，晚上也比较稳定

大家看下目前我用的各个Coding Plan推理云的反应速度。一句话，讯飞、百炼快，火山垃圾。

@Nohomework0 同感，目前只用来当搜索引擎用

都不要给Gemini充钱，Gemini现在拉到连国产大模型都不如

@GergelyOrosz I think it's more about token usage, 20 USD just can't handle it

Confirmed that Anthropic - as of now - has removed Claude Code from new Pro signups. This is what the pricing page looks like. Feels like Anthropic has the bet that those doing coding work will be willing and ready to pay at least $100/month, going forward.

@dingyi 买了跟没买一样，还算阿里的靠谱

字节的 coding plan 上次我推荐了，拿了不少佣金，现在完全花不出去，他家太慢了，慢到怀疑人生。。。服务器是不是放在了火星🤣

@luoleiorg 试了那么多coding plan， 阿里云是最稳定的，且行且珍惜吧，以后就算没了也打算买阿里的token plan了，其他平台都是垃圾。。。

智谱刚刚下手砍了老套餐权益了，昨天给阿里云 Coding Plan 提前续了一个月，在纠结要不要先续费一年，卡个位 🥹

@rmrf100 @luoleiorg 火山的没法用，一直429

@luoleiorg 火山 Coding Plan 的要好一些，有glm-5.1和 kimi-k2.6用，但是阿里云的 Coding Plan还没有