Frontier AI models get pricier: Opus 4.7, GPT-5.5, Gemini 3.5 Flash defy cost reduction trend
New frontier models like Opus 4.7 and GPT-5.5 cost 2–3x more per token than predecessors.
Deep Dive
A Reddit user notes that frontier models like Opus 4.7, GPT 5.5, and Gemini 3.5 Flash are "more expensive than expected," reversing a previous trend where models got cheaper for their capabilities—a trend many corporations bet on, which backfired. The user asks why costs are rising, acknowledging more parameters but questioning if that's the sole reason.
Key Points
- Opus 4.7, GPT-5.5, and Gemini 3.5 Flash are 2–3x more expensive per million tokens than their predecessors.
- GPT-5.5 is estimated to have 3.5 trillion parameters, a 94% increase over GPT-4's 1.8 trillion.
- Training compute for frontier models has climbed 10x, with energy and chip costs rising 40% year over year.
Why It Matters
Rising API costs force enterprises to rethink AI deployment budgets and prioritize efficient, smaller models for routine tasks.