# GPT Image 1 GPT Image 1 (model id `gpt-image-1`) is [[OpenAI]]'s first native multimodal image generation model. Released in **March 2025** as part of GPT-4o's native image generation, it replaced [[DALL-E]] 3 inside [[ChatGPT]] and shipped via the OpenAI API in April 2025. It is the **bridge model** between the DALL-E lineage and the modern [[ChatGPT Images 2.0]] surface. The architectural change was the headline; image generation moved from a separate diffusion model (DALL-E) to a single autoregressive multimodal model that handled text, vision, and image generation through the same weights. That made image generation a *capability of the assistant* rather than a separate API call. ## What it shipped - **Native multimodal generation**; the same model that handled chat now produced images, with shared context across modalities. - **Stronger instruction following** than DALL-E 3; better adherence to long, layered prompts. - **Text rendering inside images** at usable quality for the first time in OpenAI's lineup. - **Editing and inpainting** via natural-language follow-ups in the same conversation. - **API access** via the `gpt-image-1` model identifier (April 2025 release). ## Position in the lineage ``` DALL-E (2021) → DALL-E 2 (2022) → DALL-E 3 (2023) → GPT Image 1 (March 2025) → ChatGPT Images 2.0 (April 2026) ``` GPT Image 1 was the inflection point. It did not just iterate on DALL-E; it changed the architecture so that subsequent image work could integrate with reasoning, tool use, and the rest of the GPT-4o / GPT-5 stack. Without it, the [[OpenAI Product Consolidation]] pattern (DALL-E folded into ChatGPT, [[Sora]] folded into ChatGPT) would not have been possible. ## Status Superseded by [[ChatGPT Images 2.0]] in April 2026 inside ChatGPT and via the API. The `gpt-image-1` endpoint was likely retained for API consumers during the transition; check the live OpenAI models page before relying on either name in production. ## References - Announcement (4o image generation): https://openai.com/index/introducing-4o-image-generation/ - Model docs: https://platform.openai.com/docs/models - Wikipedia (GPT-4o): https://en.wikipedia.org/wiki/GPT-4o ## Related - [[OpenAI]] - [[ChatGPT]] - [[ChatGPT Images 2.0]] - [[DALL-E]] - [[GPT4]] - [[GPT-5]] - [[OpenAI Product Consolidation]] - [[AI image generation models]] - [[AI Image Generation (MoC)]]