Coda

@jyothiwrites @lucaxyzz 5.2k

@lucaxyzz he is already at 10.5K, it will blow up, maybe cause of your tweet

CEO of hermes agent only has 1500 followers 🥲

@Alibaba_Qwen Qwen 3.7 35B when?

👏👏 Introducing Qwen3.7-Plus — a multimodal agent model that unifies vision and language into one versatile agent foundation. ✅ Multimodal interactive hybrid agent: unified GUI & CLI operation across visual and text tasks ✅ Versatile coding agent & productivity assistant with full-modality input ✅ Visual Agent: perception, reasoning, grounding, and search-augmented QA ✅ Cross-harness generalization across diverse agent frameworks One model. Sees, thinks, codes, acts.🙌🙌 Now available via API on Alibaba Cloud Model Studio. Try it — let us know what you build.😎 🔗🔗⬇️⬇️ Blog：qwen.ai/blog?id=qwen3.… Qwen Studio：chat.qwen.ai/?models=qwen3.… API：modelstudio.console.alibabacloud.com/ap-southeast-1…

@MiniMax_AI Another benchmaxxed model with perfect scores that's frustrating in-use. Tbh I don't care about benches anymore.

Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities - Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas - MiniMax Sparse Attention scales context to 1M - Natively Multimodal from Step Zero API: platform.minimax.io Token Plan: platform.minimax.io/subscribe/toke… 🚀New! MiniMax Code: code.minimax.io Weights & Tech Report in ~10 Days

@ajeetprssingh Uni student plan?

got cursor pro for free for 12 months

@BrentLynch Grok video is actually good yeah. But it's strange that they invest more into image/video generation than their LLM models. Is this because X is a media app?

GROK 1.5 BEATS SEEDANCE 2.0? Let's do our own test shall we? Image to Video 720p I will say this is an incredibly difficult prompt so good job xAI on the improvements you have made definitely a noticeable improvement over 1.0. Look forward to seeing 2.0! Prompt in below thread Seedance 2.0 - Imagine 1.5 @xai

@Angaisb_ We're getting close to it.

OpenAI should release a 3D gen model so Codex can help develop full games From code to all kinds of assets

@scaling01 Yeah not hyped at all

all of this slop vagueposting only to release the most mid product in human history

@jun_song Qwen 3.7 Max is so good. But in this case, it's probably a sketchy service that gives the base URL.

Opus 4.8 distilled Alibaba Qwen 😂 The table has turned to Open Source AI

@kimmonismus No way. Anthropic will release Sonnet 4.8 first.

What?! Opus 4.8 incoming?! Holy

@kimmonismus Nah we still have around 2 weeks I guess.

Hold on, Anthropic and OpenAI releases incoming? No way

@sudoingX agreed

the best model on a single 3090 is still qwen 27B dense. nothing comes close.

Anthropic must really find ways to make their models cheaper. GPT-5.5 is at least as good as Opus 4.7 and is almost half the price. ​There was a time when the only acceptable coding agents were Anthropic models. Like 1–1.5 years ago, I was only happy with Anthropic. But now, I'm satisfied with even OSS Chinese models at 1/10th the price. So, they must release a model like Haiku 5 with a maximum price of $3/M output, while maintaining at least Opus 4.7 performance.

This is insane. With DeepSeek V4 and Qwen3.7 Max permanently discounted at $0.87/1M output tokens, @AnthropicAI has to do something with pricing.

@adonis_singh And this ranking is very legit.

gpt-5.5 at #11, wow. i do not have any words

@opencode Oh that’s great news! Qwen 3.7 is a very decent model, but the only option was the API. Now I’ll subscribe to Go again.

Qwen3.7 Max now available in Go - text only - 1M context - smartest model in the Qwen family to date

@alibaba_cloud I second this! Qwen 3.7 shouldn't be dismissed as just another Chinese model that caught up to Opus 4.7 through benchmaxxing. That spot is well deserved. I've been coding with Qwen 3.7 all day today, and its technical depth in planning actually felt superior to Opus 4.7.

Qwen3.7-Max is officially the #2 AI coding model globally. Scoring 1541 on Code Arena, it trails only Claude. Built for production: runs 35-hour tasks, 1000+ tool calls, and ships 2-week projects in hours.

@bradmillscan Really? I was feeling complete opposite x.com/i/status/20555…

Hermes responds so fast compared to OpenClaw I thought it was broken. Insanely fast.

@JSFILMZ0412 Gen AI is literally democratizing Hollywood. A movie project that costs hundreds of millions of dollars today will only cost a couple thousand and look at least similarly good.

How much would this cost to make traditionally? Seedance 2.0 — 1 Minute Long Take Part 2. WIP Multiple 15-second clips stitched into one continuous cinematic car scene. Apparently a lot of people already know the sauce… so let’s start a trend. Show me your best AI long take below 👇🔥

@thdxr Big Pickle was Deepseek v4 flash?

this is the best model i've ever used

@kimmonismus Amodei strikes me as someone who tries hard to look like a humanist, but is actually even more self-serving than Altman.

Anthropic co-founder Dario Amodei has been saying this for over a year now. And he keeps saying it. Louder each time. In May 2025, he told Axios that AI could eliminate 50% of all entry-level white-collar jobs within five years and push unemployment to 10-20%. In January 2026, he published a 20,000-word essay calling AI “a general labor substitute for humans” that will cause “unusually painful” disruption. At Davos, he warned of a “zeroth world country” forming in Silicon Valley, decoupled from the rest of society, running at 50% GDP growth while everyone else faces mass joblessness. In his own words: “We, as the producers of this technology, have a duty and an obligation to be honest about what is coming.” And the data is starting to back him up. Tech entry-level hiring dropped 30-50% in 2025. Wall Street banks are cutting ~200,000 roles concentrated at the junior level. S&P 500 companies shed employees in net terms for the first time since 2016. Anthropic’s own labor market research confirmed that 77% of businesses use Claude to automate tasks, not to augment workers. Now another Anthropic co-founder is echoing the same message: “There is a real possibility that AI will displace human labor at a very large scale. Supporting those people will be a moral imperative of historic proportions.” This is no longer a warning from the sidelines. This is the company building the technology telling you, repeatedly, that the disruption is real, it’s fast, and society is not ready for it.

@kimmonismus Would love to see the perf/mm² comparison against where EUV would have put them.

Chinas Huawei can't access EUV. So they wrote their own scaling law. The leverage of US export controls erodes. Huawei just presented the Tau (τ) Scaling Law at IEEE ISCAS, a framework that replaces geometric transistor scaling with time-based optimization across devices, circuits, chips, and systems. Huawei is not trying to win the nanometer race anymore. They're redefining it. 381 chips designed and mass-produced over six years. Kirin chips with their new LogicFolding architecture ship this fall. Target by 2031: transistor density equivalent to 1.4nm processes - without EUV (ASML embargo still hits hard) Whether Tau Scaling delivers on that promise remains to be seen. But it shows one thing: US export controls cut Huawei off from cutting-edge lithography, and instead of hitting a wall, they built a parallel road. China's semiconductor independence isn't hypothetical anymore. It's shipping in phones, running AI workloads, and now has its own scaling law presented at one of the world's top circuit design conferences. If Huawei can close the performance gap through architecture instead of lithography, the entire leverage of US export controls erodes. The sanctions were designed to keep China two generations behind. Huawei is trying to make that metric irrelevant. h/t @AndrewCurran_ for finding this interesting blogarticle and @zephyr_z9 for the photo