OpenAI Unveils GPT-5.3-Codex, Marking Major Advances in Coding and Security

Deep News
Feb 06

OpenAI CEO Sam Altman has announced the official release of GPT-5.3-Codex. The new model achieves significant breakthroughs in coding performance, with scores of 57% on the SWE-Bench Pro evaluation, 76% on TerminalBench 2.0, and 64% on OSWorld.

It introduces support for real-time control and dynamic updates during task execution. Operational efficiency has also seen a substantial leap; for completing equivalent tasks, token usage is reduced by more than half compared to the 5.2-Codex version, while single-token processing speed has increased by over 25%.

The model possesses mature computer operation capabilities. Altman also stated that this is their first model to receive a "high-level" rating in the cybersecurity dimension within the safety protection framework. Currently, they are piloting a trusted access framework and investing $10 million in API credits to accelerate the development of cybersecurity defense systems.

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Most Discussed

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10