OpenAI upgrades ChatGPT Images with faster tools

OpenAI has launched a new version of ChatGPT Images, powered by its latest image generation model, with the rollout beginning December 16 for all ChatGPT users and API users worldwide.

The model, available in the API as GPT Image 1.5, generates images up to 4x faster and offers more precise editing while retaining original details such as lighting and composition.

A new Images feature within ChatGPT allows users to generate and edit images through the sidebar, featuring preset filters and trending prompts.

Business and enterprise access is expected to follow at a later date.

OpenAI said the update improves instruction following and text rendering, and allows users to reuse their uploaded likeness across creations.

The earlier version of ChatGPT Images will remain available as a custom GPT.

🔗 Source: OpenAI

🧠 Food for thought

Implications, context, and why it matters.

OpenAI’s pricing and quota details for GPT Image 1.5 remain undisclosed, complicating adoption decisions

As of Dec 16, OpenAI’s official API pricing pages and model documentation do not list pricing or rate limits for GPT Image 1.5 1.
OpenAI charges per-token (units of text) for text models, and per-image for generation services. Costs vary by complexity or output resolution. GPT Image 1.5’s setup is unknown. Developers cannot estimate spend or compare with Midjourney or Stability AI.
Without transparent usage quotas or announced batch pricing for GPT Image 1.5, teams at large companies planning high volume image workflows cannot assess fit with budget or technical needs. Many will wait for commercial terms before integration.

Third-party providers can build Coalition for Content Provenance and Authenticity (C2PA) signing tools if OpenAI’s native content provenance support proves insufficient

OpenAI joined the C2PA Steering Committee in May 2024. It attaches Content Credentials metadata (provenance tags that state how and when content was created or edited) to DALL·E 3 outputs in ChatGPT 2. The API applies the same metadata.
Companies that need advanced compliance features such as batch watermarking, custom metadata fields, or integration with Digital Asset Management systems may find OpenAI’s native C2PA implementation too basic for regulatory or partner platform requirements.
Specialist vendors can build signing proxies, Software Development Kits (SDKs), or middleware. These tools intercept GPT Image 1.5 outputs and apply richer C2PA metadata. They can route signed assets into existing content workflows. This covers gaps for regulated fields such as media and advertising.