OpenAI added picture era performance to ChatGPT in March 2025, and the AI system created greater than 700 million completely different photographs within the first week alone.
Pushed by the recognition of the brand new characteristic, OpenAI added ChatGPT’s picture era functionality to its API by way of the gpt-image-1 mannequin. Builders and expertise professionals can now combine ChatGPT’s picture era capabilities in their very own creations.
How companies are already utilizing the gpt-image-1 mannequin
Some companies and organizations have already begun utilizing the gpt-image-1 mannequin. These embrace:
- Adobe: Sure apps, corresponding to Categorical and Firefly, give Adobe creators fast entry to the gpt-image-1 mannequin.
- Figma: A model new integration with gpt-image-1 lets customers create and edit photographs, add or take away components, and make different changes as wanted.
- HeyGen: An AI-driven video creation software, HeyGen has carried out gpt-image-1 to enhance and improve the avatar creation facet of its platform.
- Quora: This social web site with a large on-line neighborhood not too long ago made gpt-image-1 its default picture era mannequin.
Different corporations are nonetheless within the experimental part with gpt-image-1.
- Canva is exploring methods to implement the picture era mannequin into its Canva AI and Magic Studio platforms.
- GoDaddy plans to make use of the expertise to make it simpler for its customers to make logos.
- HubSpot is experimenting with AI-generated advertising collateral to be used by its prospects.
Built-in security controls
The gpt-image-1 mannequin consists of the identical built-in security controls which can be seen in ChatGPT’s 4o picture era performance; as such, the mannequin is restricted from creating dangerous imagery. Builders can alter security settings to be roughly restrictive primarily based on their software wants.
Pricing info
As with related AI fashions, the gpt-image-1 mannequin is priced per token. It consists of tiers for textual content enter, picture enter, and picture output processes.
- Textual content enter (immediate textual content): $5 for 1 million tokens
- Picture enter (enter photographs): $10 for 1 million tokens
- Picture output (generate photographs): $40 for 1 million tokens
When damaged down individually, this quantities to roughly $0.02, $0.07, and $0.19, respectively, for every generated picture. Nonetheless, the precise value varies primarily based on picture high quality and dimension.
Taking advantage of gpt-image-1
OpenAI’s gpt-image-1 mannequin is offered now. It’s usable by people and groups across the globe, though some could have to confirm their group with OpenAI to get began. Nonetheless, gpt-image-1 is already making headlines for its versatility, usability, and affordability.