OpenAI is preparing updates for its platform, testing new features related to image generation and API security. Soon, an “image model selector” may appear in the image section, allowing users to choose different generation models. It is expected that among the options will be “image-1-mini” – a new model for creating images with lower cost and extended editing capabilities. This will be useful for businesses and individual creators who need such tools.
The new selector may eventually receive the same model that operates in the Sora 2 video service, allowing the combination of image and video creation. The appearance of “image-1-mini” in the code has not yet been confirmed, but rumors about the update are growing, as the current model has not changed for a long time. Users hope to see both an improved and budget version for their tasks.
Simultaneously, OpenAI is implementing a “Guardrails” section in the platform’s navigation. This tool allows configuring data filtering through the API, detecting fabricated responses, attempts to hack the system, or change the purpose of requests. Such features are important for companies and developers, especially those working under European GDPR regulations, as it will help comply with security and transparency standards.
“Guardrails” will also be integrated into the upcoming “Agent Builder.” If data on the effectiveness of this tool becomes available, organizations will be able to better assess the quality and reliability of OpenAI models. Both new capabilities are planned to be implemented soon in the image generation sections and API settings, which will be useful for creative users and those working with the platform in work environments.
