We’ve just discovered that the LobeHub GitHub organization has been flagged and is currently hidden from public view. We haven’t received a clear reason for the restriction and believe this may be a mistake. LobeHub is an open-source project with 82K+ GitHub stars, and our development and community activities remain active as usual. @github , could you please help review our organization and restore public access? We’ll keep everyone updated here.

🚨 GPT-6 Astra is rolling out. And LobeHub is getting ready. Astra is built for the stuff that actually matters: computer use, coding, long-running agent workflows, and real professional work. As soon as broad access opens, we’re bringing it straight to LobeHub. But you don’t even have to wait. You can already connect Codex + your ChatGPT subscription directly to LobeHub today. Your models. Your tools. Your agents. One workspace. Your AI team is about to get a very serious upgrade. ⚡️

Gemini 3.8 Flash and Qwen3.8-Max-0902 are now available on LobeHub. ⚡ Gemini 3.8 Flash → ~74% on DeepSWE v1.1, up from ~65% for 3.7 Flash → 1M context → Built for long-horizon coding, autonomous agents, and complex reasoning → $0.75/M input, $3.75/M output 🚀 Qwen3.8-Max-0902 → 2.4T parameters → 1M context → Further post-trained for Coding & Cowork → Stronger long-horizon development and multi-tool workflows → $2/M input, $6/M output Both are ready to try on LobeHub.

Your Agent team just got three new superpowers. ⚡ → Files and outputs in Resources can now be sent directly to your connected chat platforms → Agent Appearance Studio is live, with custom looks, styles, and personalities → Claude Code & Codex can now run on LobeHub credits or your own provider API key The last one is still experimental. Enable Agent Provider Binding in Labs to try it. Small update. Pretty useful one.

Claude Fable 5.1 is now live on LobeHub, with a massive jump in agentic performance. 🔥 → 52.6% on Terminal Bench Science, more than 2× Fable 5 → 55.8% on Terminal Bench 4.0, up from 42.0% → Stronger across coding, research, computer use, and business workflows → Up to 45% lower cost on highly agentic workloads Built for long running tasks, deep research, and complex agent workflows. Try Claude Fable 5.1 on LobeHub today.

Ever wished you could stay in bed, head out for a walk, or even go on vacation while your Agent keeps working on the files and projects on your computer? 👀 Now you can. LobeHub Agents can work across devices. → Ask about a report stored on your computer from your phone → Edit a local project while you’re away from your desk → Let your Agent open, update, and save files directly on another device Your files stay where they are. Your Agent goes where the work is. Phone in your hand. Computer somewhere else. Work still moving. Enjoy your life🍻

🚨 GLM-5.3-Flash just landed on LobeHub, and the architecture is kind of insane. It has roughly the same total parameter count as GLM-4.5: 355B → 320B total params But under the hood: → Active params: 32B → 18B → Layers: 92 → 45 → 30T-token multimodal pretraining corpus → 1M context window → Native multimodality + tool calling Basically, @Zai_org cut the compute almost in half and somehow made the model stronger. That’s why GLM-5.3-Flash can deliver frontier-level capability at a dramatically lower cost. This is what the next model war looks like: not just bigger models, but much more intelligence per dollar. Now available on LobeHub. ⚡

🚨 Ox Alpha is about to reveal its identity. Since going live, usage has exploded 2,000x in just a few days, already surpassing V4 Flash on LobeHub. 🔥 And there’s a reason why: → Strong coding performance → Long-horizon agentic tasks → A massive 1M-token context window Still a mystery model for now. Not for much longer. When Ox Alpha is officially revealed, it’ll be available on LobeHub right away.

We may have overdone this one. 👀 LobeHub v2.2.14 just landed. This release, we: → opened Workspace beta + added Agent and Agent Group transfers → added custom names, avatars, and backgrounds for agents → brought notifications to IM channels + improved the Notification Center → added a Claude Code usage calendar → added Pi Agent + Qoder support → shipped API Key scopes, SDK, and docs → added GLM-5.3, Gemini 3.7 Flash, and Grok 4.6 Yeah, this one’s packed. There’s a lot to play with in this one.

Paste a product link. Get a finished ad video. 👀 That’s it. Our Product Ad Video Agent takes it from there: → pulls product images + key selling points → figures out the creative direction → writes the video prompts → generates the assets → runs the video model → delivers a ready-to-use ad Veo, Kling, Seedance and more. One link in. One finished video out. The Agent handles the workflow in between, end to end. Live on LobeHub. 🎬

Your AI team just got a glow-up🪄🤖 Agent Profiles are now way more customizable. 🎨Pick your vibe 🧑‍🎤Create custom avatars 🕺Go full-body ✏️Hand-drawn, anime, pixel art, Lobe style & more ✨Or just let AI cook Give every agent its own face, style, and personality. Make the researcher look nerdy. Make the coder look sleep-deprived. Make the PM look suspiciously well-rested. Same team. Completely different energy. Go make your AI crew weird👀

Ox Alpha just landed on LobeHub 🐂 ✅ 1M context ✅ Image + video input ✅ Built for coding and agentic work ✅ Free for everyone It’s already putting up some seriously impressive numbers on coding benchmarks. And somehow, we still don’t know who made it. Go give it a spin.

Three fresh models just dropped on LobeHub 👀 ✅GLM-5.3 ✅Gemini 3.7 Flash ✅Grok 4.6 Quick vibe check: GLM-5.3 → China’s latest open-source SOTA, with a big post-training upgrade over GLM-5.2. Gemini 3.7 Flash → a serious jump from 3.6 Flash. Already looking very good for frontend + game dev. Grok 4.6 → same base as 4.5, but noticeably sharper writing and responses. All live now. Pick one. Build something weird.

LobeHub v2.2.13 is live 🚀 Your agents can now: → Live inside WeChat → Generate images in any conversation → Run on Kimi K3 Fast → Connect directly to your ChatGPT subscription Less switching. More doing. All inside LobeHub.

5,000+ builders, researchers, and founders are coming to Berkeley. So are we. We’re excited to join the Agentic AI Summit 2026, alongside researchers, builders, and teams working across agent infrastructure, frameworks, evaluation, and real-world deployment. At LobeHub, we’re building an open-source workspace where people can create, organize, and collaborate with AI agents. From multi-agent teams and shared context to long-term memory, tools, skills, scheduled tasks, and human oversight, we’re exploring what it takes to move agents beyond isolated chatbots and turn them into systems that can actually work together. If you’re attending the Summit, come meet the LobeHub team, see what we’ve been building, and talk with us about the future of agentic software. 📅 August 1–2, 2026 📍 UC Berkeley 🏢 Kerr Lobby, MLK 3/F 🔎 Between Ludo Robotics and Hosted by @BerkeleyRDI See you in Berkeley. #AgenticAISummit#