OpenAI's 4o Model Image Generation Now Supports Custom GPTs, Enhancing Personalized AI Creation

AIbase基地

Published inAI News · 6 min read · Apr 21, 2025

OpenAI recently announced that its latest 4o model's image generation capabilities will support custom GPTs, providing users with a more flexible and personalized AI creation experience. According to AIbase, this update allows developers and users to build customized GPTs based on the 4o model, generating high-quality image content tailored to specific needs. The news has sparked widespread discussion within the AI community, marking another significant advancement by OpenAI in the field of personalized AI tools. Technical details have been released on the official OpenAI platform, and full user access is expected soon.

Key Highlights: Personalized and High-Quality Results

AIbase analysis reveals that the combination of the 4o model's image generation capabilities and custom GPTs brings the following breakthroughs:

Highly Personalized: Users can define unique generation styles, themes, or visual elements through custom GPTs, meeting personalized creative needs, such as generating brand-exclusive images or artwork in specific artistic styles.

High-Fidelity Images: The 4o model is renowned for its powerful multi-modal capabilities. The generated images excel in detail, color, and composition, rivaling professional design tools.

Seamless Integration: The new feature is directly embedded into the GPTs ecosystem, allowing users to enable image generation within their existing workflows without complex configurations.

Community feedback indicates that this update is considered a major upgrade to the GPTs platform since its launch. Previously, GPTs primarily focused on text processing and task automation; the addition of image generation capabilities significantly expands its application scenarios.

Wide-ranging Applications: Empowering All Scenarios from Creativity to Commerce

AIbase has compiled potential applications for 4o image generation supporting custom GPTs:

Content Creation: Bloggers and designers can quickly generate social media images, covers, or illustrations, improving content production efficiency.

Brand Marketing: Businesses can customize exclusive GPTs to generate visual materials aligned with their brand identity, strengthening their market competitiveness.

Education and Training: Educators can use custom GPTs to generate teaching charts or presentation images, enriching classroom content.

Games and Entertainment: Developers can generate game assets or concept art, accelerating creative prototyping.

It's worth noting that the 4o model's multi-language support and high contextual understanding capabilities make it widely applicable in global markets, especially suitable for scenarios requiring cross-cultural visual content.

Technical Threshold: User-Friendly, Empowering Developers

AIbase understands that the deployment of the 4o image generation feature is extremely user-friendly. Ordinary users can access custom GPTs directly through OpenAI's ChatGPT interface or API, generating images by simply inputting text prompts. Developers can utilize OpenAI's provided SDK and documentation to fine-tune exclusive GPTs based on the 4o model to meet complex needs. Hardware-wise, the functionality primarily relies on cloud computing, eliminating the need for high-performance local devices and lowering the barrier to entry. OpenAI also plans to release more tutorials and templates to help users get started quickly.

Community Response and Future Outlook

Since the announcement, the open-source community and AI professionals have shown strong interest in the 4o image generation supporting custom GPTs. AIbase has observed that some developers have already shared early test results on forums, claiming that the generation quality is comparable to professional tools like MidJourney. The community has also expressed some expectations, such as support for higher resolutions, 3D image generation, or integration with AR/VR platforms. AIbase believes that as OpenAI continues to optimize the GPTs ecosystem, more multi-modal features, such as video generation or audio-image interaction, may be introduced in the future, further enriching the personalized AI experience.

Saying 'Thank You' to ChatGPT Costs OpenAI Millions

In today's AI-driven world, AI has become an integral part of our daily conversations and problem-solving. However, what you might not know is the significant cost associated with these seemingly simple interactions. Sam Altman, CEO of OpenAI, revealed that the company spends tens of millions of dollars annually just to process user pleasantries and polite language. Simple expressions like "thank you" and "please," while enhancing the emotional connection, represent a substantial financial burden.

iFlytek's StarFire X1 Receives Major Upgrade: Aims to Rival OpenAI in AI

On April 21st, iFlytek officially announced a significant upgrade to its AI model, StarFire X1, aiming to compete with OpenAI's models in intelligent reasoning and multi-tasking capabilities. This domestically-trained large language model excels in various general tasks, including mathematics, programming, logical reasoning, text generation, language understanding, and knowledge question answering. This upgrade incorporates data from more complex scenarios, significantly improving the model's performance.

OpenAI CEO Critiques Politeness as Wasteful in AI Interactions

OpenAI CEO Sam Altman recently argued that using polite language like "please" and "thank you" with chatbots like ChatGPT is wasteful, consuming excessive electricity and computing resources. Altman suggests that while politeness may be culturally expected or perceived to improve interaction quality, it unnecessarily burdens AI systems. Each instance of polite language adds to the computational load.

OpenAI's o3 Model Test Scores Questioned; Actual Performance Falls Far Short of Claims

OpenAI's recently released o3 AI model has sparked controversy over its benchmark test performance. While OpenAI confidently claimed in December that the model could correctly answer over a quarter of the highly challenging FrontierMath math problems, this assertion starkly contrasts with recent independent test results. The Epoch Institute's independent testing revealed the model achieved only a 10% success rate, significantly lower than advertised.

OpenAI Releases 34-Page Guide to Building Intelligent Agents: From Web Search to Code Generation

On April 17, 2025, OpenAI announced via social media the release of a 34-page "Practical Guide to Intelligent Agents," offering developers comprehensive guidance on building agent applications. This marks another significant advancement in OpenAI's efforts to promote the practical application and standardization of AI technology. According to the announcement, the guide details how to utilize OpenAI's Responses API to build agents capable of web search, file search, and computer interaction. The Responses API is an evolution of Cha...

OpenAI's Stargate Project Expands Internationally, Targeting Europe

OpenAI's "Stargate Project," a joint initiative with Oracle and SoftBank, is reportedly considering international expansion, specifically targeting the UK, Germany, and France. Initially designed to bolster US artificial infrastructure with a total budget of $500 billion, the project is now rumored to be taking steps towards global markets. Image note: Image generated by AI, licensed through Midjourney. The Stargate Project is currently evaluating the feasibility of overseas expansion.