Hacker Newsnew | past | comments | ask | show | jobs | submit | rohanm93's commentslogin

This is shockingly cheap for a near frontier model. This is insane.

For context, for an agent we're working on, we're using 5-mini, which is $2/1m tokens. This is $0.30/1m tokens. And it's Opus 4.6 level - this can't be real.

I am uncomfortable about sending user data which may contain PII to their servers in China so I won't be using this as appealing as it sounds. I need this to come to a US-hosted environment at an equivalent price.

Hosting this on my own + renting GPUs is much more expensive than DeepSeek's quoted price, so not an option.


> I am uncomfortable about sending user data which may contain PII to their servers in China

As a European I feel deeply uncomfortable about sending data to US companies where I know for sure that the government has access to it.

I also feel uncomfortable sending it to China.

If you'd asked me ten years ago which one made me more uncomfortable. China.

But now I'm not so sure, in fact I'm starting to lean towards the US as being the major risk.


The chances of my bank account getting hacked due to the PLA backdoor in Deepseek is higher than the CIA backdoor in OpenAI.


Right now Im much more worried about sending data to the US and A.. At least theres a less chanse it will be missused against -me-


OpenAI released a privacy filter model https://openai.com/index/introducing-openai-privacy-filter . It is small enough to run locally, I would pass my request through it first before sending to any apis.


> For context, for an agent we're working on, we're using 5-mini, which is $2/1m tokens. This is $0.30/1m tokens. And it's Opus 4.6 level - this can't be real.

It's doesn't seem all that out there compared to the other Chinese model price/performance? Kimi2.6 is cheaper even than this, and is pretty close in performance


Kimi is indeed somewhat cheap for frontier-level intelligence, but still is $4-5 per mm tokens. Deep Seek is at least an order of magnitude cheaper.


Oh, right you are. I misread where the decimal place was in the Deepseek pricing. That is incredibly cheap


Since it's open weights it'll be available on AWS Bedrock soon(ish), likely at a higher price than the official API but still coming in under those GPT-5-mini prices.


Interesting, thanks. I'll keep an eye out.


Well done. Buying this.

I love the landing page - ugh just so perfect for the HN audience. I am pretty happy with the dock but after reading your landing, I felt like I need it.

Also, don't listen to people about pricing. $10/year as you have it is already cheap considering I'll end up using it every day (if I like it). People will never be satisfied.


Had an identical reaction to that $100m email. I decided to try it again with the browser extension. My verdict is it is better than ChatGPT Atlas for the agent mode, so I see use cases for it.

But I am still surprised it's at $100mm ARR. I had thought the company had died after their initial hyped launch and didn't see anyone talking about or using the company at all since then...and we play around with a lot of AI tools. I wonder who their customers are.


I think FUD is working on Meta coming from their archrival Bytedance.

Bytedance recently launched a similar product in China and caused quite a stir, local phone brands are jostling for partnerships, the AI agent phone.

Meta probably don't want to miss the next thing, even if it turned out to be a dud.


I curated the locations like I said in the post so it's a good first-time experience, but you can use the search to literally go anywhere!


Thanks dude!


Thank you so much! Means a ton.

Totally agree - the mobile experience need a ton of work, I'll try to figure out a better UX.


I love this so much haha.

"I can only ask my employee 20 smart things this week for $20?! And they get dumber (gpt-4o) after that? Not worth it!"


I covered this in a recent issue of a career newsletter I write [1]. Some snippets from it that I think would help new joiners (not only those joining a new industry):

1. No-one’s going to go out of their way to guide you. Esp at smaller companies, the fact that you’ve been hired means you’re joining a time-poor company, and there likely won’t be a formal onboarding process. So...you’ve got to onboard yourself.

2. Adopt a mindset of personal responsibility. This isn’t going to be a passive process, but an active one.

3. Don't meet everyone for the sake of it. The goal isn't to set-up endless meetings and become a burden. Limit yourself to a few, and make the most out of them. Make your ask clear when you ask someone for their time — e.g. 'I'd like context on X and what's been done so far'

4. Get some small wins on the scoreboard. Although your priority is to learn the key skills of the job, you also want to make a great impression during your first 60 days by showing you’ve made an impact. To do that, identify some smaller projects you can knock out to get a few small wins under your belt.

5. Be ‘seen’. Make your presence known. And communicate upwards so people know you’re busy. How? Tell your manager, “Can I send you a list of 5 things I think I need to hit the ground running? I’ll then set up a 1-1 and we could go through that.”

the full essay's here [2] (and plug: the newsletter I write is called Coached. If occasional career strategy is your thing, feel free to check it out)

[1] https://coached.com

[2] https://careersupplement.beehiiv.com/p/cs233-onboarding-hard...


Your point 1 has been my experience and I DO NOT understand why the most upvoted post in this post is saying to have one on ones with every team member on the team and also to have one of ones with related teams. I can't think of a time when that has ever been able to happen, especially on outer teams. Yes, it is ideal. But most teams and people feel like they are too busy all the time and also I have found that if you ask too many questions, sometimes that can be used against you if you don't know the team dynamics at play.

Your point 3 validates that as well.


This is also sound advice for someone wanting to join an established open source project.

Pluck some of the low-hanging fruit so they can get to know you and try not to be a burden on the core maintainers. And...have fun?


I think starting off with the impression of being a small wins person sticks, and people will think of you as "does small tasks well." No promos and whatnot.

It's why being a loud jackass that deploys something horrendous has better outcomes for your career.


in Dubai and whatever they’re doing this definitely working - over just the last month-ish we’ve had a 4-8 days of extremely heavy rains (and gov alerts to stay in), when just two years ago it’d be a surprise if it rained 3 days all year


That's actually climate change and the desert is prone for that. I don't think they are seeding these heavy rains as usually they can turn into a disaster.


So good - I came across your updates on Twitter and was blown away by how mature this product is already, especially at launch. Congratulations!

I’m also a user of your other Chrome app you developed (Intention) [1] - I’d recommend people try that out too (simple free app for productivity), it’s also so good.

Good luck with the launch!

[1] https://chrome.google.com/webstore/detail/intention-stop-min...


Thanks Rohan! Glad you're enjoying Intention and thanks for checking out Browserflow. :)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: