Serhii

@trq212 @DhravyaShah is supermemory better? And how

We've rolled out a new auto-memory feature. Claude now remembers what it learns across sessions — your project context, debugging patterns, preferred approaches — and recalls it later without you having to write anything down.

@trq212 What do you use to create demo video?

@Alibaba_Qwen Distillation is crazy

🚀 Introducing the Qwen 3.5 Medium Model Series Qwen3.5-Flash · Qwen3.5-35B-A3B · Qwen3.5-122B-A10B · Qwen3.5-27B ✨ More intelligence, less compute. • Qwen3.5-35B-A3B now surpasses Qwen3-235B-A22B-2507 and Qwen3-VL-235B-A22B — a reminder that better architecture, data quality, and RL can move intelligence forward, not just bigger parameter counts. • Qwen3.5-122B-A10B and 27B continue narrowing the gap between medium-sized and frontier models — especially in more complex agent scenarios. • Qwen3.5-Flash is the hosted production version aligned with 35B-A3B, featuring: – 1M context length by default – Official built-in tools 🔗 Hugging Face: huggingface.co/collections/Qw… 🔗 ModelScope: modelscope.cn/collections/Qw… 🔗 Qwen3.5-Flash API: modelstudio.console.alibabacloud.com/ap-southeast-1… Try in Qwen Chat 👇 Flash: chat.qwen.ai/?models=qwen3.… 27B: chat.qwen.ai/?models=qwen3.… 35B-A3B: chat.qwen.ai/?models=qwen3.… 122B-A10B: chat.qwen.ai/?models=qwen3.… Would love to hear what you build with it.

@gabemulley Great to hear because I switched to playwright MCP because /chrome was too slow for quick UI iterations

We've been working hard to make Claude in Chrome as fast as possible. We are rolling out an experimental "Quick Mode" to all users today. It is particularly impressive when combined a fast model like Opus 4.6 (fast mode) or Haiku 4.5. I look forward to hearing what you think!

@AnthropicAI That’s sad. It means the only way to get cheap models is to train huge ones. Good for $NVDA tho

We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.

@GoogleDeepMind And yet Gemini 3 Pro is still in preview and has crazy inconsistency

Gemini 3.1 Pro is here. We’ve significantly improved the model’s overall intelligence so it can solve tougher problems. 🧵

@MCamblor97 @sama Lol makes no sense. He was the best ad campaign Anthropic ever had

@soholev @sama They were focused in threatening Peter for the name he used when the project went viral. Wrong vision 🥴

Peter Steinberger is joining OpenAI to drive the next generation of personal agents. He is a genius with a lot of amazing ideas about the future of very smart agents interacting with each other to do very useful things for people. We expect this will quickly become core to our product offerings. OpenClaw will live in a foundation as an open source project that OpenAI will continue to support. The future is going to be extremely multi-agent and it's important to us to support open source as part of that.

@diminunez @sama Can you give more context? I missed that

@soholev @sama They started sending legal love letters to peter

@OpenAI I could tell it’s OpenAI ad in the first 5 seconds

You can just build things.

@NPH_On_X @adocomplete 200k as before

@adocomplete @soholev What is the current context window in general for a regular chat? And, what will this mean for users?

It's been a busy day and I hope you're enjoying Opus 4.6. By the way, we're offering a limited-time $50 (USD, or local currency equivalent) in extra usage to Pro and Max users to coincide with the launch today.

@melvynx U have no tests? Most models can build whatever you need as long as you create reliable tests and just ralph it

Claude Opus 4.6 is out. Try for 4 hours to refactor my entire Codebase to Convex. +1000 files edited The results? Nothing works.

@oikon48 @adocomplete Oh.. I am on Max 20x subscription

@adocomplete @soholev I could with API

@adocomplete How to enable 1M context?

New for Claude Opus 4.6 - Adjustable effort levels Effort levels control Opus 4.6’s adaptive reasoning, which dynamically allocates thinking based on task complexity. Lower effort is faster and cheaper, while higher effort provides deeper reasoning for complex problems.

@OpenAIDevs @OpenAI I assume SF? Need some events in NYC

Doors are open at the Codex hackathon 🛠️ We're excited to be here with the Codex community and will be sharing behind-the-scenes clips and updates in this thread throughout the day. You can just build things.

@AnthropicAI “Mostly”? Meaning you had to click allow every 2 minutes?

@drummatick I use a Claude Code in terminal inside Cursor lol

Somehow they convinced everyone that coding in a CLI is better than using an IDE along with the agent It’s strange how Twitter hive mind works

@zaidmukaddam With Claude’s pricing it’s game over only to my wallet…

FYI it has 1M context window it's game over now

@trq212 Is focus now shifting more towards everyday people from coding nerds? Cowork, law plugins, excel and power point plugins.

Opus 4.6 is a special model, it really feels like a true collaborator you might have got a sneak peek at its work earlier this week- the videos I launched were made completely by Opus 4.6 (see below)