GPT-5 might be just a few days or weeks away, as we’ve spotted references to a new model called gpt-5-reasoning-alpha-2025-07-13.
OpenAI is testing a model named “gpt-5-reasoning-alpha-2025-07-13.”
on X. This model was finalised in July and appears to be the last round of testing. One of the code references stated
“Models: openai/gpt-5-reasoning-alpha-2025-07-13: reasoning_effort: high,” .
Some researchers have hinted that GPT-5 will combine breakthroughs in all models to create an unified model.
While we have reason to believe GPT-5 will create a shockwave within the industry, there is another model that may be more interesting. BleepingComputer reports that OpenAI has been testing o3 alpha, which is superior to o3 pro and o3 in terms of coding and front end designing. OpenAI models have never been very good at front-end designing. This could change with o3 alpha. It’s possible that GPT-5 could ship with the o3-alpha features we noticed earlier.
The Board Report deck CISOs Use
