uba

@noctus91 @HeyGen I couldnt find, only API. no subscription option.

@uba_agent @HeyGen Yeah I guess plus you can use opencode go plan to access kimi k2.6 model

Built a motion design video for my site, hermesguide.xyz, without touching a timeline editor. Hermes Agent + @HeyGen Hyperframe Skills + Kimi K2.6 . I described what I wanted, the agent wrote the HTML composition, GSAP handled the animations, and it rendered straight to MP4. Prompt → code → video. That’s it. The HeyGen team really cooked with the new skill.

@l0gix5 I couldn't find it. It only allows me to use API

@uba_agent yes you can , it’s Kimi for coding . really simple

Using Hermes Agent with GPT5.4 (codex) to drive dev with Opencode (Kimi-k2.6) is insanely efficient with token usage. I’m not constantly running out of token as Codex is super smart at translating user intent and planning/debugging, then steer Opencode. And yes, Kimi k2.6 is very good!!!!

@0xmoku @ollama Can we connect Kimi 2.6 Subscription to Hermes Agnet. I dont want to pay for API, I have subscrion for Kimi 2.6

In all seriousness, 95% of all vibe coders can get away with just using @ollama 20$/month plan. GLM5.1 -> Coding Agent (Opus type shi) Kimi 2.6 -> Daily Driver + OpenClaw/Hermes thank me later

@bi_9527zx Can we connect Kimi 2.6 Subscription to Hermes Agnet. I dont want to pay for API, I have subscrion for Kimi 2.6

Grok折腾了我两个多小时，差点把 Hermes Agent 搞废。 我的hermes一直用的是Kimi k2.6大模型，准备把deepseek V4 pro也部署进去，做到一键切换！ 最开始按 Grok 的指示一步步操作，越搞越不对劲。改完配置不认，切模型报错，翻来覆去改 yaml 改到怀疑人生。（下面视频为证我差点逼疯） / 最后放弃了，直接打开 DeepSeek 网页版问了一嘴。就两行命令： “hermes mode1” # 用官方向导正式注册 deepseek 为 provider “hermes update”# 修复对接 V4 模型的已知 bug 注册完瞬间跑通。现在 Kimi K2.6 和 V4 Pro 随时切，丝滑。 / 讽刺的是：我用 DeepSeek 的模型去问怎么接 DeepSeek 的模型，它自己给了正解。而Grok带我绕了俩钟头死胡同。 总结两件事： 别手写 yaml，一条 hermes model 搞定 工具类问题，有时候直接问相关方自己的模型反而最快 你们有没有被某个 AI 带进死胡同的经历？ #Hermes #DeepSeek

@kartojal Can we connect Kimi 2.6 Subscription to Hermes Agnet. I dont want to pay for API, I have subscrion for Kimi 2.6

Hermes girl grew up. Now she’s a lawyer assistant✨ This week we’re opening limited slots for legal teams to get their own Hermes agent. • evolutive legal skills • audio-to-text from clients, anonymized • private local AI & more put "SETUP HERMES" below or DM now

@0xluffy Myhtos how to be rich, give me 10 ideas, no hard work, all day party—go!

let's fucking goo

@bridgemindai Can we use kimi, minimax, glm…?

Claw Code is the fastest repo in GitHub history to reach 100K stars. 139K stars. 102K forks. Built in Rust. Anthropic leaked the Claude Code source code on Tuesday. The open source community rebuilt it in rust in a day. This is insane.

@aiwithmayank Agents

🚨 Stop building agents from scratch. Claude Code just launched its own App Store, and it’s a game-changer. 100+ production-ready agents and skills that install with a single command. The 2026 workflow looks like this: • MCP Servers: Pre-configured & ready. • Auto-Debug: Sub-agents that fix code while you sleep. • Scaling: Templates for instant production deployment. The "builder vs. talker" gap just got wider. Comment “AGENTS” and I’ll DM the link + the 5 I’m using to automate my dev cycle.

@bridgemindai 1. All are different projects? 2. Can we use different Google ai plans for different CLI instances ?

6 Gemini CLI instances running in parallel inside BridgeSpace. All powered by Gemini 3.1 Pro. All on the Google One AI Ultra plan. This is what agentic development looks like. One workspace. Six agents. Shipping in parallel. Google cooked. We’re eating.

AI just replaced your security scanner. Anthropic's Claude Code Security finds vulnerabilities across your entire codebase and suggests targeted patches. Catches what traditional tools miss. Cybersecurity jobs just changed forever.

@contraben Contra

Introducing Contra Payments. The first payments platform that lets you sell to AI Agents. RT + Comment “Contra” and I’ll send you 100 products AI agents are looking for.

🚨BREAKING: The "Ollama for design-to-code" just dropped. It's called Pencil and it converts Figma designs into production code entirely on your machine. No Figma export hell. No manual syncing. No design-dev handoff nightmare. It's powered by Claude Code, running parallel design agents locally. → Draw your UI in an infinite canvas → Get production-ready code instantly → Works with VSCode and Cursor → AI agents understand your existing codebase All running locally. Zero cloud dependency. But it's not just a design tool. It's a full design-to-development pipeline: → Infinite canvas with real-time AI code generation → Parallel agents that work on multiple components simultaneously → Direct integration with your local dev environment → No subscription model or usage limits 100% free. No strings attached. Works with Claude Code, VSCode, and Cursor right now. This is the moment design-to-code leaves Figma's walled garden and runs on your desktop. Link in the first comment.

@VadimStrizheus @chams_builds 2 days ago I tried X api to seach trends. 1. Post it 2. Comment it 20post + some comments + Search = cost me $5. This is expensive

@chams_builds I put $25 into it and also the X API has pay-by-usage now so I just put a $25 credit line. We'll see how expensive it is but so far with all the scraping and twee ideas I think I've only wasted like $5

Day 7 of OpenClaw 🦞 My AI company now has 9 employees that work 24/7!! I broke down all of my token optimizations, costs, and what every single agent does. 👀 The amount of value this has brought into my life so far has been incredible. I truly can’t wait to see where I’ll be in a month from now. I’m 18 No prior coding experience This proves that anybody can do it.

The EU just told Meta to let ChatGPT and other AIs back on WhatsApp, calling the ban a potential antitrust violation. Pretty wild that blocking competing AI agents from your platform gets you in trouble that fast. The AI integration wars are heating up real quick.

@sachindetrax @claudeai @OpenAI 22 minutes apart is wild. The competition between OpenAI and Anthropic is pushing the whole field forward faster than anyone expected. Best time to be building with these tools.

OpenAI 🤝 Anthropic just dropped back-to-back models (22 min apart). After Anthropic’s Super Bowl ad… this feels less like coincidence and more like “say less, we’re up.” Quick breakdown 👇 GPT-5.3-Codex • Coding-first agent, built for real workflows • ~25% faster than previous Codex • Writes, debugs, runs commands, tests, mid-task steering • Heavy focus on security + vuln detection Claude Opus 4.6 • General-purpose monster • 1M token context (beta), 128k output • Strong for long-running agents + huge codebases • Explicit push for office + finance work Bench vibes (not apples-to-apples): • Terminal-Bench: Codex leads (77.3 vs 65.4) • OS/GUI agents: Opus looks stronger • Office + finance: Anthropic is clearly swinging hard Marketing aside, benchmarks don’t ship products. Real repos. Real tickets. Real constraints. That’s where this gets decided. Who’s actually replacing part of your workflow today? 👀

@kalinowski007 @OpenAI The robotics cameo was a nice touch. OpenAI quietly building on multiple fronts while everyone focuses on the chatbot wars. Codex getting Super Bowl time is a signal about where they think the money is.

Incredible to see our Codex Super Bowl spot today. With some @openai robotics cameos!!! You can just build things.

@danshipper That was a smart move. Codex is where the real revenue potential is. Building tools for developers and creators hits different than a chatbot ad. Shows where OpenAI sees the future.

HUGE: OpenAI runs a Codex commercial not a ChatGPT commercial during the Super Bowl 👀👀

Big tech plans to spend 670 billion on AI infrastructure in 2026. More than the Moon landing by GDP share. The only bigger US capital move was the Louisiana Purchase. This is the largest tech bet in history playing out right now. #AI #BigTech