OpenRouter and Artificial Analysis evaluated AI model performance and cost.
They tracked intelligence scores and prompt pricing per million tokens.
This analysis covered various categories over a 30-day period in late 2025.
Weekly token usage data from OpenRouter informed this broader assessment.
Market leadership shifted frequently among model creators due to new releases.
Examples
- Programming applications demand high intelligence, scoring 0.51.
- Science applications also require high intelligence, with a score of 0.47.
- SEO applications pay the most for tokens, averaging $1.40 per million.
- Legal and translation applications pay significantly less, around $0.16-$0.17.
- Customer service chatbots prioritize speed and low cost for high-volume queries.
It's easy to get caught up in the weekly token shifts when new AI models drop, making it seem like the market is a constant leadership battle for supremacy.
But thinking those quick jumps automatically tell us about a model's true intelligence or its actual market price feels like missing the full story arc.
My friend, Leo, might grab a cheap new AI for a one-off project, which spikes its usage for a week, but then he quickly pivots back when he finds it just can't handle the nuance his users need.
That 30-day assessment tries to get at the deeper why, but if it's just raw numbers, it still might not capture the real-world choices developers make, where short-term cost savings can lead to long-term headaches.