OpenAI's o1 Pro Mode Released! Is it really WORTH it?

[@the_dented_NPC]

Bro spent $200 for us🤗

[@MervinPraison]

🙏

[@sebkeccu4546]

Yeah, that's crazy expensive. 2400 dollars per year. And according to the small letters, the unlimited is not really unlimited (there is a star next to it, referring to the fair use policy) I also think you don't have priority when systems are impacted, it does not have a segregated tier like the enterprise subscription. With the enterprise subscription, your catgut continues to work even when the public catgut is down or making errors, because they are separate instances. At 2400 dollars per year, they should also have their own SLAW guarantees. I also think that in the first month there might be many subscribers like this channel that purchase it to try it for just one month, This will artificially inflate the numbers, but it will drop hard next month when those temporary subscriptions fall. You should also know that the pro version cannot be combined with the team subscription...therefore you will pay double. The team subscription was bad anyway, because it acts like drugs, if you don't renew it every year, they delete ALL your chats and ALL your BOTS. They could have blocked you from sending out new queries, but deleting existing data is just bad. Especially because you can't migrate them to the personal, plus or pro subscriptions... you just lose everything. And that is going to happen now in January, because that is when it was introduced last year. I've disabled the auto-renewal of my team subscription, otherwise I can't switch over to the more advanced models like pro for another year.

[@Matthias_Kuehn]

These are all examples either from the video of openai or exactly from other KZbinrs. Did you actually purchase the pro plan? Please provide some unique tests in the next video

[@zalmaytech]

your videos are exceptional. I casually came across your channel and really enjoy your perspective and evaluations. Subscribed!

[@Corteum]

Can it be used to build an web app or mobile app?

[@MyExRebirth]

Also, as with the C# Three Sum problem, both GPT4o and 1o (Not Pro) generated the correct code as soon as I gave them the code template and test cases.

[@mohammedhasanin2611]

can you confirm whether its possible to upload pdf now and do web search?

[@MyExRebirth]

I asked the Gpt-4o and 1o the trolley problem and both solved it without issues, plus o1 provided a long explanation. I don't know why your attempt got a different response?

[@jw-sx4pi]

is it a separate model from the 01 that we can use with normal pro license or is it o1 but with access to more compute?

[@MervinPraison]

o1 with more compute and reasoning capability = o1 pro mode

[@itstime8777]

​@@MervinPraison Basicially the amount of reasoning time is increased ?! Is the model able to output better results in the same time-amount?

[@sadmansakib5715]

Can anyone tell me if there is a prompt limit in O1 Pro mode?

[@florianl6764]

Please do a comparison between o1 mini, o1 and o1 pro

[@Vishnu-jq4yj]

Hi Mervin, ❤ from TamilNadu. What do u think is the nearest competitor to this o1 pro model, among open source models..also I would love watching a comparison. Thank you

[@sebkeccu4546]

There are many, they are all comming out in the next few months. Because its not difficult to achieve. It's just the old model but with reiterations. ANd some models have already passed 4o model, which means that the other thinking models might be better. But we will see, because it might be that in the next few days maybe there will be a upgrade for 4o. Something else to consider: opensource llm's are getting really close to the 4o performance. Chechkout qwen 2.5, Claude sonnet 3.5 v2, Llama 3.3 and especially Mistral is a special one. Mistral does't make much publicity, but according to the benchmarks it is at the same level as chatgpt and even exceeding it at some benchmarks. They announced a o1 pro alternative comming out in january, hopefully that one will be affordable (and more stable, because chatgpt has many downtimes lately, it's victim of its success and not able to grow fast enough). Mistral feels way faster because of that. Personally I now use a lot of small llm's on my phone and laptop such as llama 3.2 and QWen 2.5 7b (the VL version) . I don't know how they made it possible for such a small vl language to be so good at recognising the real world. At his moment google is out of the picture, the results are behind openai, mistral, cloud, qwen and llama. Even the german Aleph Alpha is better.

[@sebkeccu4546]

SO if I was you, I would go for claude sonnet 3.5 v2(new) its better then all the openai models, faster and much cheaper. And it is expected to get its upgrade in the comming weeks.

[@ভয়ানক-ত৪জ]

Hello, now will O1-PREVIEW Model and O1 Model be able to be selected separately? Or is there no O1-PREVIEW?

[@vlonethug0]

no preview now

[@imanister3625]

Only o1 model now

[@sebkeccu4546]

The o1 preview was better than the o1, because it had more processing time. They didn't want it to compete with the o1 pro, therefore they had to remove it.

[@60pluscrazy]

Thanks 🎉

[@Techonsapevole]

Thanks for testing it.. but Llama3.3 70B is better for my needs and pockets :)

[@DadEncyclopedia]

I‘m not smart with GPT. How do I know if he is worth it?

[@reddddzzz]

Why are you asking questions that openai asked? of course its going to be in the dataset...

[@HyperUpscale]

It is 2 times smarter and you pay 10 times more.. I dont think thats a good deal :D

[@ah89971]

I don't see any improvement in coding though

[@haroldpierre1726]

The value proposition is that it could potentially replace an employee or make employees several times more productive. As a business owner, I am thinking about it!!

[@radektheplayer]

@@haroldpierre1726 what a greed

[@haroldpierre1726]

@@radektheplayer Or the opposite of greed. My expenses have skyrocketed over the past several years and we have increased our charges by slightly more than 3% in 8 years. We continue to be the low cost provider in our market by doing everything to increase efficiency. My competitors charge 40% more than we do.

[@user-ze9tj9yj4t]

@@radektheplayerIf you have a business you’d probably do the same

[@DadEncyclopedia]

I think the so-called thinking time is to share the resources and load of the host, not thinking at all.

[@MarkRodriguez-i1d]

Thank you so much for this amazing video! Could you help me with something unrelated: I have a SafePal wallet with USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). How should I go about transferring them to Binance?

[@zalmaytech]

these scams are everywhere

[@simtangaranvijay273]

Bro are you tamil?

[@Kanna5460]

Yes

[@simtangaranvijay273]

@@Kanna5460 how you know?

[@unclecode]

Thx for the content, and have fun with O1 Pro ;) I wonder why it couldn't get the C# example. In training data, these models have seen Python code more often, and if O1 really has a good reasoning skill, it must be able to transfer the skill from Python to C#, that's why I wonder why it couldn't handle this. And a suggestion: can you compare the same questions with Claude 3.5, while adding CoT and asking Claude to think and plan.

[@MervinPraison]

Thank you. Sure will do.

[@matheus_palandi]

Título em português vídeo em inglês Top

[@HyperUpscale]

o1 is not a model, it is an agent or even more - agents. Comparing Agents with model is like comparing team of professionals with a student....

[@irakliirakli4267]

bro spent 200$ for 260 likes 💀

[@AamirFarooq]

Please add audio track Urdu or Hindi

[@radektheplayer]

Thats the end of jobs

[@muratoztosun5859]

Not budget not for us.

[@monstrositylabs]

I'm using it now, and it's WAY worse than the preview version IMO. It's so bad I'm considering closing two of my three accounts