Hasan Can retweeted
Inkling 2 Small’s weights are now available for download on HuggingFace under an Apache 2.0 license.
Make Inkling your own!
huggingface.co/thinkingmachin…
Make Inkling your own!
huggingface.co/thinkingmachin…
91
271
1,816
283,919
Pretty good
Today, Search is getting a massive upgrade with Gemini 4 Hydrogen! It’s the biggest model improvement to Search yet. And it’s available now for all users, globally.
Why it’s awesome:
- Crazy fast
- Cutting-edge intelligence, with advanced reasoning, more precise instruction ool calling
- More visual responses, including images and custom tables, charts, and simulations
- More scannable response layout
- Better understanding of intent (including for long multi-part queries with ambiguous typos)
- Broader range of links to the web (improved reasoning -> more query fan-outs)
TL;DR: powerful and fast. Take Hydrogen for a spin in Search and let me know what you think!
0
0
1
64
Good performance
Introducing Gemini 4 Hydrogen, an extremely capable and fast model built to bring Gemini 4 to everyone!
Available to all tiers of Google AI Studio, Gemini API, Gemini for business, the Gemini app, Google Search, and paid Google AI Pro + Ultra subscribers in Google Antigravity.
0
0
0
80
Hasan Can retweeted
Jensen Huang and the Nvidia board just made a massive mistake.
He should not have resigned.
The board should have taken the time to conduct a thorough investigation, consult with shareholders, and carefully evaluate the long-term implications of this decision.
There is no clear replacement for Jensen Huang, and we believe his departure will ultimately be a net negative for Nvidia, $NVDA, and its shareholders.
This will go down as the most costly mistake of the AI Revolution.
He should not have resigned.
The board should have taken the time to conduct a thorough investigation, consult with shareholders, and carefully evaluate the long-term implications of this decision.
There is no clear replacement for Jensen Huang, and we believe his departure will ultimately be a net negative for Nvidia, $NVDA, and its shareholders.
This will go down as the most costly mistake of the AI Revolution.
1,423
2,104
25,900
1,124,494
Lmfao, what a ridiculous price. At those prices, people would rather use much better models via API. Who would even bother with Gemini 4 Flash at that price when there are so many great, affordable agentic API models out there? I doubt it'll see much use outside Gemini subs.
Gemini 4 Hydrogen is now in the SKU Catalog, likely releasing later today
Pricing:
Input 1.50$
Output: 7.50$
They might launch it at an introductory price, as they did with Carbon. 0.75$ / 3.75$
0
0
2
131
Update: DeepSeek V4.1 Flash now performs best in CommandCode for me. It’s ahead of Hermes, DSH, and every other harness I’ve tried with this model.
DeepSeek V4.1 Flash reaches its full potential inside DeepSeek Harness. In Hermes and CommandCode, even at max effort, it simply isn’t as good as it is in DSH. DSH burns a lot of tokens and tends to drag tasks out, but quality improves noticeably.
GLM 5.3 Flash still has better harness agnostic performance and cheaper.
2
0
2
167
I switched to CommandCode's GOAT plan after looking for alternatives because of OpenCode's limits. Right now, I'd say it's the provider I'm most satisfied with among those offering open-source models, especially in terms of speed, usage limits, model quality, and the harness henever I have Opus 5.5 usage available, is to use Opus as the main orchestrator, with DeepSeek V4.1 Flash 1017 as worker agents.
The idea is that Opus handles the reasoning, architecture, and overall direction. It's the brain of the operation. I give it simple instructions, and it understands my intent, delegates complex tasks to the DeepSeek workers, oversees their work, and delivers the final result.
The idea is that Opus handles the reasoning, architecture, and overall direction. It's the brain of the operation. I give it simple instructions, and it understands my intent, delegates complex tasks to the DeepSeek workers, oversees their work, and delivers the final result.
0
1
3
65
Hasan Can retweeted
Anthropic is pushing for GPUs and other hardware chips (TPUs/NPUs/etc) to not be able to run Opensource models thus ensuring Local AI is impossible
This is them wanting to stop you from using your hardware to own your intelligence
AMD & Google wanted it done, NVIDIA did not
This is them wanting to stop you from using your hardware to own your intelligence
AMD & Google wanted it done, NVIDIA did not
Anthropic now plans to push for govts to require firmware level locks on Western GPUs that prevent "unapproved" (Chinese open-source) models from being run on them, in the name of "AI safety" and "national security."
AMD/Google are on board, only Jensen opposed it
1/Community noteThis is a satirical account parodying The Daily Mail; no such information has been revealed by credible sources.
dailymail.co.uk
125
191
1,695
103,121
Hasan Can retweeted
怎么确认你的 Gpt 6.1Sol 有没有灰度到 Gpt 6.2 Sol?
网友测试 6.2 Sol 正在 以6.1Sol 的 ID 灰度测试，不联网可以说出fable的发布日期和6月份的新闻
同时进行指纹测试，用modeltrace测了下指纹结果出来个opus5.5 55% 的归因 概率（GPT 也开蒸了？）
纯玩梗 ，别当真
网友测试 6.2 Sol 正在 以6.1Sol 的 ID 灰度测试，不联网可以说出fable的发布日期和6月份的新闻
同时进行指纹测试，用modeltrace测了下指纹结果出来个opus5.5 55% 的归因 概率（GPT 也开蒸了？）
纯玩梗 ，别当真
3
1
18
5,241
Hasan Can retweeted
GPT-6.2 Astra and Sol appeared on the Quotas page in the Google Cloud Console
Release likely tomorrow
Release likely tomorrow
8
19
362
96,785
For the first time in a long while, Google has pulled this decisively ahead of its competitors, yet nobody has access to it, not even Elon Musk, one of the richest people in the world.
Can we finally stop calling model and benchmark announcements "releases" when nobody can 's the point of a model that users can't test in the real world?
Can we finally stop calling model and benchmark announcements "releases" when nobody can 's the point of a model that users can't test in the real world?
0
0
5
371
Hasan Can retweeted
Sam Altman thinks RAM sitting in personal computers is a waste when it could be put to better use in data centers.
Weird. I was just thinking that all those billions sitting in billionaires’ bank accounts could be put to much better use in our communities.
Weird. I was just thinking that all those billions sitting in billionaires’ bank accounts could be put to much better use in our communities.
442
13,288
91,068
1,023,940
Hasan Can retweeted
Plus subscribers make up 86.2% of the paying user base, while Pro users represent roughly 1%.
Yet the biggest upgrades and new features seem to keep landing on Pro first.
Most people aren't going to pay $100–$200 per month just to get a better ChatGPT experience.
Plus is where the majority of paying users are. Give us stronger usage limits and access to Dots as well.
It's time the Plus plan got the attention it deserves.
Yet the biggest upgrades and new features seem to keep landing on Pro first.
Most people aren't going to pay $100–$200 per month just to get a better ChatGPT experience.
Plus is where the majority of paying users are. Give us stronger usage limits and access to Dots as well.
It's time the Plus plan got the attention it deserves.
299
179
4,165
415,959
Hasan Can retweeted
ANTHROPIC IS CRACKING DOWN ON VIDEO GAME REWRITES 🚨
Claude Code picked up a new rule around Oct 27, sent from Anthropic’s servers under a flag called "tengu_heron_brook". There’s no setting to turn it off.
The full rule: “Do build original games, games in a genre, and games igns and assets. Don’t build a recreation of a specific existing game, even if it is renamed or called original.”
Is this the end of game rewrites, modding and everything in between?
Claude Code picked up a new rule around Oct 27, sent from Anthropic’s servers under a flag called "tengu_heron_brook". There’s no setting to turn it off.
The full rule: “Do build original games, games in a genre, and games igns and assets. Don’t build a recreation of a specific existing game, even if it is renamed or called original.”
Is this the end of game rewrites, modding and everything in between?
839
518
9,200
2,426,712
Hasan Can retweeted
Sonnet 5.5 costs 2.8x GPT-6.1 Sol per task on the API. On $20 plans, a third
12
9
383
13,736
Hasan Can retweeted
关于Gpt-6.1-terra 消耗过高的真相：
原来Gpt-6.1-terra 隐藏倍率高达 1.9x，实际订阅里只比 Gpt-6.1-sol 便宜30%。（via:linuxdo）
再加上 6.1 Terra基本上是6.1Sol一倍以上的速度，所以体感上 6.1 Terra消耗的额度就是 6.1 SOL 的 1.5 倍以上
原来Gpt-6.1-terra 隐藏倍率高达 1.9x，实际订阅里只比 Gpt-6.1-sol 便宜30%。（via:linuxdo）
再加上 6.1 Terra基本上是6.1Sol一倍以上的速度，所以体感上 6.1 Terra消耗的额度就是 6.1 SOL 的 1.5 倍以上
8
6
64
10,423
OpenAI won't have the balls to merge Codex and regular ChatGPT usage limits. Just watch.
What made ChatGPT special, what set it apart from Claude, was that even after hitting your Codex limits, you could still use ChatGPT to do research with the exact same models. You could still use it for simpler tasks.
Nobody is going to keep paying for ChatGPT if it becomes just like Claude, especially with limits like these and while it's already so far behind Claude. People will simply cancel their subscriptions and switch straight to Claude AI.
What made ChatGPT special, what set it apart from Claude, was that even after hitting your Codex limits, you could still use ChatGPT to do research with the exact same models. You could still use it for simpler tasks.
Nobody is going to keep paying for ChatGPT if it becomes just like Claude, especially with limits like these and while it's already so far behind Claude. People will simply cancel their subscriptions and switch straight to Claude AI.
0
0
6
279