OpenAI’s New AI Model Comes with Upgrades and Limited Availability
OpenAI has launched GPT-4.5, an updated AI model designed to improve creative intelligence and conversational depth. However, due to ongoing GPU shortages, access is currently limited to ChatGPT Pro subscribers, priced at $200 per month, and high-tier API users. OpenAI is also integrating Sora, its AI-powered text-to-video tool, into ChatGPT, bringing video generation directly into the chat interface.
The company has positioned GPT-4.5 as an incremental improvement over its previous models, offering better reasoning, creativity, and response coherence performance. The model has been trained with new techniques to improve contextual awareness, making its responses feel more natural. Sora integration marks a new step in AI-driven video creation, allowing users to generate short, high-quality video clips from text prompts within the ChatGPT ecosystem.
What’s Happening & Why This Matters
OpenAI’s decision to launch GPT-4.5 with restricted access highlights AI companies’ challenges in balancing innovation with infrastructure limitations. GPU shortages have made it difficult to scale AI services when demand is rising. Initially, OpenAI planned to include ChatGPT Plus users, who pay $20 per month, in the rollout. However, limited computational resources allow only Pro and API customers to use GPT-4.5 at this stage.
Sam Altman, OpenAI’s CEO, acknowledged that demand for GPT-4.5 exceeded expectations, making it difficult to provide universal access. The company is working to increase its GPU supply and expand availability to lower-tier subscribers in the coming months. OpenAI has already committed $500 billion toward AI infrastructure under its Project Stargate initiative, which aims to enhance its long-term processing capabilities.
While GPT-4.5 is currently only available to select users, the company has hinted at future accessibility improvements. As GPU availability improves, more users could gain access to the model, bringing its advanced capabilities to a wider audience.
Sora Video Integration Brings AI-Generated Video to ChatGPT
Alongside GPT-4.5, OpenAI is rolling out Sora, its AI-powered video generator, into the ChatGPT ecosystem. This update allows users to create short videos directly from text prompts within the ChatGPT interface. Currently, Sora can generate up to 20-second video clips, and OpenAI plans to expand its functionality over time.
The company has been testing Sora’s capabilities through a standalone web portal, but the upcoming ChatGPT integration will allow users to generate videos without leaving the platform. OpenAI is also developing a dedicated Sora mobile app that could expand video editing capabilities.
Rohan Sahai, OpenAI’s product lead for Sora, confirmed that the company is working on an upgraded version called Sora Turbo. This new iteration is expected to improve video quality and processing speed, making it a more viable tool for users looking to create AI-generated content efficiently.
AI with ‘Vibes’ and More Natural Interaction
One of the serious updates in GPT-4.5 is the introduction of a more human-like interaction model. OpenAI has focused on making AI conversations feel more natural by refining how the model responds to user input. This includes improvements in tone, engagement, and adaptability to different conversational styles.
The company introduced a new “vibe test” to measure how well the AI can generate responses that feel collaborative and emotionally intelligent. GPT-4.5 has been trained to rewrite messages to enhance user engagement, making conversations feel more fluid. Engineers have also worked on refining the model’s ability to guide user thinking more intuitively.
This shift moves away from rigid, AI-generated responses and creates a more dynamic conversational experience. Users interacting with GPT-4.5 will notice that it responds in a way that feels more natural, reducing the robotic tone often associated with AI interactions.
TF Summary: What’s Next?
OpenAI is actively working to resolve GPU shortages to increase access to GPT-4.5. The introduction of Sora video integration marks a new chapter in AI-driven creativity, bringing video generation into the ChatGPT ecosystem. As OpenAI refines the responsiveness and personality of its AI models, users can expect more fluid, human-like conversations. For now, top-tier customers get early access, while the rest wait for OpenAI to scale up its infrastructure and make these advancements available to a wider audience.
— Text-to-Speech (TTS) provided by gspeech