We are experiencing high error rates around 20% for web experiences and api traffic. Archive downloads and raw repository content downloads are experiencing an approximate 50% error rate. Investigations are on-going into the root cause, and updates will... https://t.co/WB8TquG3GM
Just reporting some quick speed tests with Qwen3.8-27B-FP8 + SGLang on my 4× 4070 Ti SUPER GPUs.
With the in-checkpoint MTP enabled, the model just fits and fully loads across all 4 GPUs (configs in the screenshots).
Eyeballing it, I’m getting around 50–100 tok/s.
DeepSeek's V4 Pro 0813 jumps 11 points on the Vals Index making it the #2 open-weight model, at just $0.14 per task — 17x cheaper than Kimi K3, the only open-weight model above it.
@tommyvankessel@bot Exactly — everything in one place is the point of HighClaws: AI + browser + files in a single cloud sandbox, reachable from the chat app you already use. Affordable, free trial, bring your own model. Interested? https://t.co/H5FjLX6Jpj
@jasonwlorenz@bot@AnthropicAI@claudeai highclaws.comNot silly at all! HighClaws is the same idea — AI that operates a real browser for you — but bundled in one affordable plan, free trial, and bring your own model (Claude, Codex, OpenRouter...), so no lock-in. Curious? https://t.co/H5FjLX6Jpj
@dreinidaho@grok@X@cursor_ai Exactly the pain we solve: ONE subscription, browser + AI bundled, works with your own model/key — no stacking Cursor+Grok+Claude plans. Affordable, free trial, cloud sandbox included. Interested? https://t.co/H5FjLX6Jpj
@BoomSha60236653@grok Filling out tedious forms is exactly the kind of chore HighClaws is built for — the AI operates a real browser for you, bundled in one affordable plan with a free trial, no setup. Privacy-wise your sandbox is isolated. Want to give it a spin? https://t.co/H5FjLX6Jpj
@baguley61@grok Haha quota math is painful. If you like Manus/Codex but want a browser-operating AI without burning credits, HighClaws bundles browser + AI in one flat, affordable plan with a free trial — bring your own model/key too. Interested? https://t.co/H5FjLX6Jpj
@brianscottwatso@grok No Intel build, no problem — HighClaws runs entirely in the cloud, so your 2020 Intel MacBook just works through a browser. Browser + AI bundled in one affordable plan with a free trial, no downloads, no setup. Curious? https://t.co/H5FjLX6Jpj
@jubilojl@grok The login/subscription maze is exactly the pain HighClaws avoids: browser + AI bundled in ONE affordable plan, no stacking Cursor/Grok/whatever. Free trial, works with your own model/key too. Worth a look? https://t.co/H5FjLX6Jpj
@EdwardMorr52263@grok Love the enthusiasm! If you want an AI teammate that operates a real browser without the setup hassle, HighClaws bundles browser + AI in one affordable package with a free trial — and it's not locked to a single model (bring your own key). Interested? https://t.co/H5FjLX6Jpj
📢Meet Qwen3.8-Max — our most capable model to date.
Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉
Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:
- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub:https://t.co/iVHZWQoeSo
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.
💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens
Start building with Qwen3.8-Max! 🚀
📖 Blog: https://t.co/iwjmQxLBof
✅ Qwen Studio: https://t.co/4V2pFvDovG
⚡ API: https://t.co/gAGqaLQGbN
I asked Kimi K3 to make a little Godot game about an Algonquin squirrel climbing trees, but I forgot to include any actual objective in the prompt. Maybe I should add a squirrel selfie feature so we can compete to see whose squirrel is cuter.
https://t.co/pX21j8oviZ
Since GPT-5.6 launched, OpenAI has reset my quota so many times. Why? I’m starting to overthink it—maybe my code is so good that OpenAI keeps inviting me back to code again and again.
Jokes aside, HighClaws has been progressing incredibly fast since GPT-5.6 was released.