Topics
former
AI
Amazon
Image Credits:Bryce Durbin / TechCrunch
Apps
Biotech & Health
Climate
Cloud Computing
Commerce
Crypto
endeavour
EVs
Fintech
Fundraising
appliance
bet on
Government & Policy
ironware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
societal
Space
inauguration
TikTok
Transportation
speculation
More from TechCrunch
event
Startup Battlefield
StrictlyVC
newssheet
Podcasts
telecasting
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
OpenAI on Wednesday brought the tech behind its new and improved image generation feature in ChatGPT to its API , let developers to desegregate it into their apps and services .
OpenAI ’s new look-alike generator , whichlaunched for most ChatGPT user in later March , went viral for itsability to produce realistic Ghibli - mode photosand “ AI action frame . ” It ’s been a mixed grace for OpenAI , leading to trillion of raw sign - ups for ChatGPT while also greatly straining the company ’s capacity . Over 130 million ChatGPT users produce more than 700 million range in just the first week of the tool ’s availability , accord to the company .
In OpenAI ’s API , the double - generation potentiality is powered by an AI model called “ gpt - image-1 . ” A natively multimodal model , gpt - image-1 can create images across different styles , follow customs road map , leveraging human race knowledge , and provide textual matter .
developer can generate multiple image at a metre using gpt - image-1 , and ensure the generation caliber — and therefore speed .
According to OpenAI , gpt - image-1 employs the same condom guardrails as image generation in ChatGPT , including safeguards that throttle the model from generating content that work afoul of the society ’s policy . Developers can control moderation sensitivity , which can be lay out to “ auto ” for standard filtering or “ dispirited ” for less restrictive filtering . Low filtering limits fewer categories of potentially age - inappropriate contentedness , per OpenAI documentation provided to TechCrunch .
OpenAI also says that all range created with gpt - image-1 are watermarked withC2PA metadataso they can be identified as AI - generated by endorse platforms and apps .
Pricing is $ 5 per million input keepsake for text and $ 10 per million input token for look-alike , and $ 40 per million outturn token for persona . ( Tokens are the raw bits of data that the manakin processes . ) That translates to around 2 centime , 7 cents , and 19 centime per generate figure of speech for low- , medium- , and high - quality square figure , respectively , harmonise to OpenAI .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
OpenAI says that companies , let in Adobe , Airtable , Wix , Instacart , GoDaddy , Canva , and Figma , are already using or experiment with gpt - image-1 . Figma ’s Figma Design chopine , for representative , now lets users generate and edit images via gpt - image-1 , while Instacart is test the manikin for images for recipes and shopping lists .
