Dispo Mail

@qoder_ai_ide Benchmark and capabilities?

Beyond Ultimate. Cantus is here. ( /ˈkan(t)əs/ ) A new model, now live in Qoder — above our Ultimate tier. Built for long-running agents, made for the hardest knowledge work and coding tasks. Half price from today. For a limited time.

@DurovPD Are u sure buddy?

@pierceboggan Who still uses copilot subscription bro!

GPT-5.5 is 60% for GitHub Copilot Max and 40% off for Pro+ subscribers this weekend!

@EntelligenceAI Senator, I'm Singaporean.

🚨 NEW MODEL ALERT Singapore just dropped a model that's putting up frontier-level numbers. Agnes 2.5 Pro: • 82.7 on SWE-bench Verified • 78.7 multilingual • Strong gains on SWE Atlas • Beating GLM 5.2 and DeepSeek V4 Pro on multiple cuts • Free API available today The biggest takeaway isn't the benchmark here. It's the country. NOW frontier is no longer just a US vs China story.

@whateverjessN @MissJacque_line The same thing they told about China being a decade back in AI & advanced semiconductors. LoL

@MissJacque_line Try harder !!! At least 10 years behind spaceX and mostly coping and stealing other technologies

They crushed Tesla with BYD and they gonna crush SpaceX with the price 10x lower. Go CHINA 🇨🇳

@ZackKorman That zcode will go down with the GLM/Z AI when they'll be irrelevant as other models give better results with less cost.

Zai, the Chinese lab that made GLM-5.2, has a harness (Zcode) that is identical to Codex. China really not beating the copying allegations.

@0x0SojalSec Few years after, don't get surprised when Chinese SpaceX overtakes American SpaceX. Then that'll be ur national security issue like always.

SpaceX Falcon 9 reached this milestone in 2015. China's Falcon 9 moment 10 years later.

@MrAhmadAwais @CommandCodeAI Do u guys inject ur own system prompts in requests at those endpoints or that just acts as transparent proxy and used only for billing?

@MailDispo @CommandCodeAI You can use CommandCode.ai/provider

Hey @CommandCodeAI I want to use commandcode from vscode with github copilot. Is it possible, how? The latest features "Custom Endpoint" API settings in vscode coplilt allows users to integrate third party AI providers integrate into copilot.

@MrAhmadAwais People ignoring meta/Facebook the same way zuck ignores people's private & sensitive data and his ethics to exploit them. Who in the sane brain uses meta ai! We rather pay more for other better AI models.

Why is everyone ignoring Meta's new coding model? Better than GLM 5.2 yet cheaper & it has vision. Meta's new Muse Spark 1.1 model is the new underdog, being ignored by everyone. Command Code is the only coding agent where you can actually run Muse Spark 1.1 right now. We shipped it right after the launch. Meta has entered the extremely crowded AI coding market with this model, and it's pretty cheap. In the last 24hrs, there were ZERO errors on this model. I know the usage was on the lower end for us, but not that low to have no errors at all. This means, it's probably going to be a pretty nice coding model. I've been running it for a couple hours. It's legit. I actually want them to open source it. I believe Mark Zuckerberg needs a lot more push in the developer space to make this work. It's sad to see that every single partner listed on Meta's own announcement chose to ignore this launch. Cursory search, I couldn't find any. Did I miss something? None of them offer Muse Spark 1.1 integration. So what exactly are those testimonials worth? If you're going to put your name on the announcement and then ship nothing, your quote means nothing tbf. Developers in our discord community already wondering why no one else cares. We care. We've shipped it live on day one. Go use it. If Meta's leadership is reading this, partner with people who "give a shit". Zuck needs way more real push in the developer space to make this land. It's that simple. DM's open.

@jfversluis Done. Bottom line: Give more love to Maui the same as Ms is doing with AI/Copilot. Just look at kmp/react-native with expo, Maui is years behind & still not worth developing general publicly used mobile apps. Have to hack & patch things esp UI, very rare good libs.

If you build apps with .NET MAUI, we would really appreciate a few minutes of your time. There is a .NET MAUI survey open now to help us understand what is working, what still hurts, and where we should focus next. Link below, and thank you!

@pierceboggan Want to have DeepSeek v4 in the plans. And why left vs 2026 copilot as step child?

VS Code 1.128 is out with some great improvements to the Agents window (including grouped + workspace-less chats) and bring your own key! code.visualstudio.com/updates/v1_128

@MrAhmadAwais @krishav_kr9125 Not really better than DeepSeek v4 flash. But it's nice to have it for free.

@krishav_kr9125 it's a smaller model. i'd say better than dsv4flash

Hy3 model is now available for free in Command Code. Super nice open model, Apache 2.0 licensed, hosted in the US. Available on plans. We just 4x'd the capacity. 💙 LIVE NOW 🔹cmd update or npm i -g command-code@latest 🔹Run `cmd` and `/model` select Tencent Hy3 (FREE)

@pierceboggan Copilot still not usable at all. If you remove vscode/vs tight integration, it's the useless thing ever. Make it affordable like opencode, ollamacloud, qoder. 10$ plan with against claude vaporizes with one command, while claude 20$ plan can do same 1000x times.

Back from some time off with the best present... GitHub Copilot subscriptions now include open-weight models, starting with Kimi K2.7 Code!

@PaulShellDev @pierceboggan It's Microsoft horror naming convention for decades. Get used to it.

@pierceboggan VS Code harness is separate from GitHub Copilot harness? For some reason I thought it was all the CLI underneath now. Am I hallucinating? 🤔

Optimizing models for use in the GitHub Copilot harness is a continuous process that involves deep partnership with model partners like OpenAI! Post-launch, we also run online experiments to provide continuous improvement to models. This particular experiment improved latency, token efficiency, and trajectory time for GPT-5.5!

@wieslawsoltes Bro is here to eat avalonia ui team launch. Bro took things very seriously single the last issue 😺

I just love when people use my project , in this case LibreWPF and make it run on Linux, community is the best. This is running AvalonDock on top of LibreWPF/ProGPU WebGPU rendered on Linux #issuecomment-4890690076" target="_blank" rel="nofollow noopener">github.com/Dirkster99/Ava…

@boxmining @Zai_org That's a crap piece of shit. Tell the glm to write a basic 1000 lines of code, boom 2m tokens consumed. The zcode is something we never needed, never wanted crap.

Chinese labs aren't just dropping model weights anymore, they're building full ecosystems. @Zai_org's ZCode + GLM 5.2 = native IDE, 1M context, agentic workflows, and now dSpark speculative decoding. The open model platform war is here. 👇

@GHchangelog Add deepseek, qwen series model too. Opencode is kicking ass 1000x for the same 10$ than copilot.

Kimi K2.7 Code, an open-weight model, is now generally available as a selectable option in GitHub Copilot. • Available on Pro, Pro+, Max plans with rollout starting in Visual Studio Code and other dev environments github.blog/changelog/2026…

@CommandCodeAI Claude Sonnet 5 always. GLM 5.2 isn't worth for the significant high number of tokens it consumes.

Claude Sonnet 5 or GLM 5.2 — which one?

@nainaverse How is this not a modern form of prostitution?

@jietang Make models similar to Claude for same version: opus class, sonet class. Optimize token consumption, add support for thinking effort levels, add image support, context length adjustable. Without sonet class, I can't replace Claude with GLM.

Any new features we must have in the next version of glm?