OpenAI’s O3 Price Drop Reshapes the Landscape for Vibe Coders

On June 10, OpenAI significantly reduced the prices of its o3 reasoning model by approximately 80%, dropping rates from $10 to $2 per million input tokens and $40 to $8 for output tokens. This led to immediate changes among API resellers, with Cursor equating o3 requests to GPT-4o calls and Windsurf lowering its “o3-reasoning” tier to one credit, resulting in a ten-fold reduction in costs for Cursor users. Alongside the price cut, latency improvements were noted, with o3 feeling more responsive due to new Nvidia GB200 clusters and an upgraded scheduler, though it remains slower than lighter models. In contrast, while Claude 4 boasts impressive speed and a large context window, users report issues with its reliability. It tends to create fictional functions and inaccurate unit tests, leading to frustration despite its fast performance.

Source link