OpenAI has commenced the rollout of its latest AI model, GPT-4.5, to users subscribed to the ChatGPT Plus tier.
According to a series of updates on X, OpenAI indicated that the rollout will span “1-3 days” and anticipated changes in rate limits. GPT-4.5 was initially released for subscribers of OpenAI’s $200-a-month ChatGPT Pro plan last week.
“We aim to provide all users with access to GPT-4.5 a substantial rate limit, although we anticipate that these limits may evolve as we gather more insights about demand,” the company stated.
GPT-4.5 represents OpenAI’s most substantial AI model to date, developed with more computational power and data than any prior releases. However, it doesn’t necessarily rank as OpenAI’s top performer—in several AI benchmarks, GPT-4.5 falls behind newer reasoning models from DeepSeek, Anthropic, and even OpenAI itself.
Running GPT-4.5 is also notably costly, according to OpenAI. The company is currently contemplating whether to provision GPT-4.5 in its API for the long haul due to these expenses. As a consequence, GPT-4.5 has become the most expensive model available in OpenAI’s API, with pricing set at $75 per million tokens (~750,000 words) input and $150 per million tokens output, amounting to 30 times the input cost and 15 times the output cost of OpenAI’s widely-used GPT-4o model.
Nonetheless, OpenAI asserts that the increased scale of GPT-4.5 has endowed it with “greater world knowledge” and “enhanced emotional intelligence.” The model is also reported to have fewer hallucinations compared to many others, theoretically making it less likely to produce inaccurate information.
Moreover, GPT-4.5 exhibits strong persuasive capabilities. An internal benchmark revealed that the model excels at convincing other AIs to provide it with funding.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
