March 2, 2012 - Technology media outlet The Decoder published a blog post yesterday (March 1) reporting that OpenAI Latest Models GPT-4.5 With limited performance gains, the cost has increased dramatically, raising questions in the industry about its price/performance ratio.
According to the official performance comparison data provided by OpenAI, the improvement of GPT-4.5 over GPT-4o is 6.8-13.2 percentage points:
- PROFESSIONAL QUERY: GPT-4.5 outperforms GPT-4o at 63.2%.
- DAILY QUERY: GPT-4.5 is superior to GPT-4o in the case of 57%.
- Creative task: GPT-4.5 outperformed GPT-4o at 56.8%.
However, GPT-4.5 is much more expensive than GPT-4o, costing $75 per million tokens for GPT-4.5 versus $2.5 for GPT-4o on the input side, and $150 per million tokens for GPT-4.5 versus $10 for GPT-4o on the output side.This means that GPT-4.5 has 30 times the input cost of GPT-4o and 15 times the output cost.
OpenAI has made it clear that GPT-4.5 is not a replacement for GPT-4o and is evaluating the feasibility of making it available via APIs over the long term.OpenAI is actively collecting feedback from users to assess whether GPT-4.5 has enough unique value to justify its operating costs. If GPT-4.5 is eventually discontinued, developers who build applications specifically based on the model will face challenges.
OpenAI may release an optimized version of GPT-4.5 Turbo to reduce costs. However, given the limited improvements in GPT-4.5, it is doubtful that the Turbo version will offer significant advantages. Previously, the GPT-4 Turbo was optimized for speed and price, but typically fell short of the original GPT-4 in terms of quality.