OpenAI's expensive GPT-4.5 might vanish from the API, leaving developers in limbo
OpenAI's latest language model GPT-4.5 might not remain available through the API long-term, which could create problems for developers who have built their applications specifically around this model's capabilities.
According to OpenAI, GPT-4.5 costs substantially more than GPT-4o due to its massive size and computational requirements. The company has explicitly stated it isn't meant to replace GPT-4o. At $75 per million input tokens and $150 per million output tokens, GPT-4.5 is dramatically more expensive than alternatives like GPT-4o ($2.50/$10) or o1 ($15/$60).
OpenAI already questioned the long-term API availability of its particularly powerful but expensive model when it was first announced, saying it needs to balance supporting current capabilities with developing future models.
It remains unclear whether GPT-4.5 truly offers unique capabilities that other models like GPT-4o can't provide. OpenAI is actively seeking user feedback about whether GPT-4.5 delivers unique value for specific use cases. Developers who build applications exclusively around GPT-4.5 might face challenges if the model is discontinued.
Is a GPT-4.5 Turbo coming - and is it worth it?
One possibility is that OpenAI might release an optimized, more cost-effective version, perhaps called "GPT-4.5 Turbo." The company has used similar strategies with previous models like GPT-4 and GPT-3.5.
But given that GPT-4.5 already offers only modest improvements over GPT-4o, it's questionable whether a streamlined version would offer meaningful advantages over other models from OpenAI or its competitors. While previous turbo versions of GPT-4 did offer faster speeds and lower prices, they were typically inferior to the original GPT-4 in terms of quality, although this is a heated debate and there are people who prefer the turbo models.
When comparing GPT-4.5 and GPT-4o directly, the performance gap appears quite narrow: GPT-4.5 outperforms GPT-4o in just 63.2 percent of professional queries, 57 percent of everyday queries, and only 56.8 percent of creative tasks. This raises questions about GPT-4.5's cost-effectiveness for most use cases.
AI News Without the Hype – Curated by Humans
As a THE DECODER subscriber, you get ad-free reading, our weekly AI newsletter, the exclusive "AI Radar" Frontier Report 6× per year, access to comments, and our complete archive.
Subscribe nowAI news without the hype
Curated by humans.
- Over 20 percent launch discount.
- Read without distractions – no Google ads.
- Access to comments and community discussions.
- Weekly AI newsletter.
- 6 times a year: “AI Radar” – deep dives on key AI topics.
- Up to 25 % off on KI Pro online events.
- Access to our full ten-year archive.
- Get the latest AI news from The Decoder.