Price per 1M tokens is meaningless
4 hours ago
- #Token Efficiency
- #Model Comparison
- #AI Cost
- Comparing AI model costs by price per 1M tokens is meaningless due to differences in tokenizers and token efficiency.
- Each frontier lab uses its own tokenizer, causing the same text to be split into varying token counts, making direct price comparisons unreliable.
- Token efficiency—how much is achieved per token—varies widely, especially with 'thinking' tokens in chain-of-thought processes, significantly impacting overall costs.
- A benchmark table shows models like GPT-5.5 can have lower cost per task despite higher token prices, while cheaper per-token models like GLM-5.2 may be less token-efficient.
- DeepSeek V4 Pro stands out for extremely low cost per task despite lower intelligence scores, highlighting cost-efficiency disparities.