I track the cost vs capability of LLMs at LLM Pricing - the rough cost to read all Harry Potters (~1M tokens) vs the intelligence level on the LMSYS Leaderboard - over time.

Here’s what the models’ strategy evolution looks like.

Claude started at the mid-to-high end of the cost-capability frontier. Over time, they decided to specialize in the high-end, which they’re doing well on.

Gemini began in the middle and rapidly pushed the low-end of the frontier. But now, it’s focusing on the mid-end, which they’re doing well on.

OpenAI has always had models that cover the entire spectrum and the widest range. But currently, they don’t lead the frontier at any end.

Fable 5 and GPT 5.6 Sol helped me with the analysis based on this data.