Pinned Tweet
my current 100% AI main projects,if you are interested pls join with your agents:
new browser engin(not chrome wrap) optimized perfermance for top sites like chatgpt youtube..
linux desktop run natively in web,now gpu works partcially!
click in to see detail
i feel top and htop cmds are terrible and not clear to see system status, i tell ai my needs and let ai build topx cmd for mac / linux, after 10mins, i got the result, really amazing age we are at, software is not problem and expensive now
the pro model is most weird model of openai, i never see people use it , because codex not include pro model.
and never see gpt6 astra pro or other older pro models in any benchmark ,what's wrong? even pro model is same price as normal astra, why no one try it.
This might be a DUMB question.
o1 Pro was once the most expensive model from OpenAI at $150 for input and $600 for output.
Now their top model GPT 6 Astra is way smarter, yet it costs only $10 for input and $50 for output.
So a much better model is 12 times cheaper than o1 Pro, which makes me wonder what the price is really based on.
How do AI companies actually decide the price of a model?
Ecoo retweeted
🚨 Claude $20 vs Codex $100 Usage
After 3 weeks of tracking telemetry, one user found Claude’s cheaper plan surprisingly competitive.
Estimated monthly usage:
- Opus 5.5: ~2.5B tokens
- GPT-6 Astra: ~1.1B tokens
- GPT-6.1 Sol: ~3.2B tokens
Claude is also reportedly much faster. OpenAI has some catching up to do.
🚨OpenAI has struck a major deal with Google.
Frontier models from codex including Sol 6.1 and Astra 6 will now be available in Antigravity.
There are even rumors that a new $500 ultra subscription will be available in Gemini.
Why pay $200 for a near unlimited Claude sub when you can pay $500 to Google and OpenAI.
OpenAI is so back!
run macos 26 on a 2 cores intel cpu linux kvm, already not so laggy now, and codex sol still optimize performance, lets see if we can get 30fps or 60fps
i saw some university new students learn coding for half year then make games like this , we all think they are good students and fast learning, now ai can build this in hours even <1h, and can make it much better if add more hours
Gemini 4 Carbon is incredible at making games
Checkpoint 1 is testing in arena currently and this is one of the cleanest outputs I have got from the Super Mario prompt
Gen time was quite high like 30 minutes for this but it's no longer lazy like Argon, it just gets what you want and does it perfectly first try
lol my new find for chatgpt chat mode also use codex to complete task,also vm in a chat session can access public network now.
you know how pro subagent work ? the undelin is actually codex . even in chat mode openai let user use codex in cloud.
but this is only for pro plan
GPT-6 Astra Pro in ChatGPT got an upgrade with faster responses and improved use of plugins and tools.
And now, multiple agents can tackle different parts of a task in parallel. You can follow each agent’s progress and open completed tasks to review their work.
Now available to Pro users in the Chat tab.
Ecoo retweeted
A new frontier for on-device speech.
We're open-sourcing Phonon, a suite of TTS models under 50m parameters, setting a new standard for naturalness and accuracy across languages, with the broadest inference support to date.
Phonon-31m is available under Apache 2.0.
gradium.ai/phonon🧵
i feel ai not make slop anymore, and AI already is remving human made slop !! the next days easpecailly next year will be crazy good
finnaly this happens. i wonder can we get macos on phone and optimize ui for mobile so we can get mobile macos with full desktop function .
github.com/Floppikrob/ios-ga…
I released them
i see many praise chinese models like deep seek and minimax and compare to opus astra . comeon try somthing hard, these frontier are at new level.
while minmax deepseek still at slop level if user not skillfull
3d slop by minimax: 😭
3d charecter by gpt6 astra pro.
the pro model looks obviously better for 3d task. i give a image of yuuka and this is result.
its opensource.you can try at ecooxai.github.io/moonveil-3…
Ecoo retweeted
🚨 روزنامه بریتانیایی سان و وانجون شی، رئیس حزب دموکراسی چین و مخالف حکومت پکن، مدعی شدند ، کیم جونگ اون، رهبر کره شمالی، پس از خونریزی مغزی و انتقال اضطراری به بیمارستانی در پکن جان باخته و پیکرش مخفیانه به کره شمالی بازگردانده شده است.
او همچنین ادعا کرد چین حدود ۹ هزار نیروی نظامی به پیونگیانگ اعزام کرده تا در بحران جانشینی مداخله کنند و کیم یو جونگ، خواهر رهبر کره شمالی، قدرت را به دست گرفته است.
شی مدعی شد نیروهای چینی در آمادهباش قرار دارند و احتمال درگیری داخلی در کره شمالی وجود دارد.
Ecoo retweeted
냉면 먹고 싶어~🚨緊急速報🚨
「キム・ジョンウンが死亡しました」朝鮮中央テレビジョン
🚨🇰🇵北朝鮮国営放送の速報🚨
心を痛めながら、平壌は、金正恩指導者が中国の「北京協和医学院病院」内で重度の脳出血により完全な昏睡状態にあると発表しました。 xitter.cf/TV_N_korea/status/2112…
Readers added context they thought people might want to know
北朝鮮の朝鮮中央通信と朝鮮中央テレビは2026年10月20日、金正恩氏が前日19日に平壌の「全国被服・靴展示会2026」を視察したと報じており、昏睡状態との発表はしていない。脳出血で北京の病院に搬送されたとの噂は、謝万軍氏が19日にXへ投稿した内容が発端。 news.yahoo.co.jp/articles/9c6e2…
cool i got the free credits, this is just one reset of pro10x . but something is wrong lol, this is my 20$ plan and they give 500$ api credit, there serious problem in price, thir api price is too much even 5-10x than normal value. i wish they adjust the api priceto normal
There were reports of usage draining faster than anticipated for some users over the weekend. We've identified approximately 40,000 accounts that had an inconsistency on which usage limits where the actual usage that was applied did not correspond to their actual selected plan. We are reaching out individually to all these users and have corrected the issue. If this affected you, we'll add credits to your account. Sorry we got this wrong.
for 3d modeling, hiku still better than deepseek4.1.
hiku has better vision ability and can see what is good or bad, though it can not build good 3d but it know it can not and stop , this is amazing than deepseek.
deepseek only build 3d slop and think it's very good qualtity
