Ddos 
OpenAI has recently unveiled a new addition to its lineup of reasoning models—o3-pro—designed to harness greater computational power, enabling deeper contemplation and consistently more precise responses.
Demonstrating exceptional performance in mathematics, science, and programming, o3-pro is set to replace the current o1-pro model within ChatGPT. However, access is presently limited to subscribers of the $200/month ChatGPT Pro plan and users of ChatGPT Team.
According to OpenAI, the o3-pro model is equipped with tool invocation capabilities to help users make the most of its potential. Depending on user prompts, o3-pro can perform web searches, analyze documents, reason over visual inputs, execute Python code, and utilize memory for personalized responses.
In comparison, the o1-pro model lacks access to tools, making it functionally inferior to o3-pro. Yet o3-pro is not without its drawbacks; its more intensive computational processes lead to longer response times, meaning users may frequently encounter delays in receiving answers.
OpenAI’s internal evaluations reveal that o3-pro is more favorably received across key domains including science, education, software development, business, and collaboration. It consistently scores higher in clarity, thoroughness, adherence to instructions, and factual accuracy.
Nevertheless, o3-pro comes with several current limitations:
-
Temporary chat memory features are disabled while OpenAI resolves technical issues
-
The model does not support image generation; users must use 4o, o3, or o4-mini for visual outputs
-
The model does not support the Canvas feature
Eligibility for o3-pro Access:
-
$200/month ChatGPT Pro subscribers
-
$25/month ChatGPT Teams subscribers
-
Coming soon: ChatGPT Enterprise subscribers
-
Coming soon: ChatGPT Education subscribers
Starting today, developers can access o3-pro via the OpenAI API platform. The model supports a 200K context window and is priced at $20 per million input tokens and $80 per million output tokens.
Due to its deliberative nature, OpenAI recommends developers use o3-pro in background mode to avoid timeouts. In this mode, API tasks will continue running even if the developer’s session is interrupted.
The model’s training data is current up to May 31, 2024.
Donate