Do $100–200/month LLM plan get access to smarter models than $10–20/month plans?
At a given point in time. It typically just gives earlier (by weeks/months) access to more powerful/specialized models, but they’re usually rolled out to the $20 tier eventually as well. Off the top of my head:
OpenAI’s $200 subscription:
Is the only way to get access to o1 Pro, which was the highest-compute reasoning variant. Currently I think it’s considered superseded in capabilities by o3, available at $20/month.
Was the only way to get access to Deep Research for, I think, ~1 month.
Was the only way to get access to Operator for some time.
Is currently the only way to get access to Codex, though it will probably become available to the $20 tier and/or via API pricing.
Will probably be the only way to get access to o3 Pro if and when it comes out, at least for some time.
Anthropic’s $100 subscription:
Is currently the only way to access their Deep Research variant.
I think o1 Pro is the only model that never became available at $20/month, and o3 Pro will maybe be the same.
Is the only way to get o1 Pro, which was the highest-compute reasoning variant. Currently I think it’s considered superseded in capabilities by o3, available at $20/month.
It sounds like you’re saying o3 is better than o1 Pro, so there’s no reason to pay extra for o1 Pro?
(I am not the first person to observe that OpenAI’s naming scheme is terrible)
It sounds like you’re saying o3 is better than o1 Pro, so there’s no reason to pay extra for o1 Pro?
I believe so. Benchmark performance is better I think, comparing here and here, and I believe the description for o1 Pro in the model-picker menu for the $200-tier users has become “former best reasoning model” or something like this after o3 came out.
At a given point in time. It typically just gives earlier (by weeks/months) access to more powerful/specialized models, but they’re usually rolled out to the $20 tier eventually as well. Off the top of my head:
OpenAI’s $200 subscription:
Is the only way to get access to o1 Pro, which was the highest-compute reasoning variant. Currently I think it’s considered superseded in capabilities by o3, available at $20/month.
Was the only way to get access to Deep Research for, I think, ~1 month.
Was the only way to get access to Operator for some time.
Is currently the only way to get access to Codex, though it will probably become available to the $20 tier and/or via API pricing.
Will probably be the only way to get access to o3 Pro if and when it comes out, at least for some time.
Anthropic’s $100 subscription:
Is currently the only way to access their Deep Research variant.
I think o1 Pro is the only model that never became available at $20/month, and o3 Pro will maybe be the same.
It sounds like you’re saying o3 is better than o1 Pro, so there’s no reason to pay extra for o1 Pro?
(I am not the first person to observe that OpenAI’s naming scheme is terrible)
I believe so. Benchmark performance is better I think, comparing here and here, and I believe the description for o1 Pro in the model-picker menu for the $200-tier users has become “former best reasoning model” or something like this after o3 came out.