Source image file
AI Models & Assistants / Knowledge library
chatgpt-image-latest
Previous image model used in ChatGPT.
Why learn chatgpt-image-latest?
Offers the documented OpenAI model capabilities through hosted.
The problem it solves
Provides model inference for supported text / image inputs.
Where it came from
chatgpt-image-latest: Previous image model used in ChatGPT. Published aliases: chatgpt-image-latest. Input: text, image. Output: image, text. Checked 2026-10-13. Model capabilities are distinct from tools available in a particular agent.
Build your understanding
Work through these exercises in order. Try the task before opening a reference.
- Explain what chatgpt-image-latest does.
- Identify a use case for chatgpt-image-latest.
- Describe a limitation of chatgpt-image-latest.
- Name a compatible tool for chatgpt-image-latest.
- Compare chatgpt-image-latest against an alternative.
Documentation and sources
- Primary source / model card (developers.openai.com)
Catalog research notes
- See the AI Workbench for source dates, pricing evidence and configuration details.
Connect the ideas
Built on
Real World Example
Legacy Asset Archive Restoration
Historical archival firms managing large collections of degraded physical photographs.
The business problem
Manual restoration of damaged historical images is labor-intensive, inconsistent across different restoration artists, and fails to scale when processing thousands of archival assets.
How this topic solves it
Utilize the chatgpt-image-latest model via the image edit endpoint to automate the inpainting of missing or damaged sections of digitized historical photographs based on textual descriptions of the surrounding context.
- Digitize the physical photograph to create a high-resolution source image.
- Define a mask layer identifying the specific damaged or missing areas requiring restoration.
- Construct a prompt describing the visual characteristics of the missing content to guide the model.
- Submit the source image, mask, and prompt to the v1/images/edits endpoint using the chatgpt-image-latest model.
- Review the generated output for historical accuracy and consistency with the original image style.
The flow shows how a damaged asset is processed through the model to produce a restored version.
Defines damaged regions
Inpainting engine
Final output
- Digitized Asset → input → Masking Tool
- Masking Tool → masked image + prompt → chatgpt-image-latest
- chatgpt-image-latest → generates → Restored Asset
Expected business outcome
A standardized, repeatable workflow that accelerates the restoration process while maintaining a consistent aesthetic across diverse archival collections.
Limitations and checks
The model may struggle with complex, non-repeating textures or highly specific historical details not present in the training data; verify results by comparing the output against original reference material to ensure fidelity.