Tokens and AI API pricing
AI APIs charge per token - small chunks of text - usually quoted in US dollars per million input tokens and per million output tokens, with output typically priced higher.
Your cost depends on how much you send (input), how much the model writes back (output) and, for reasoning models, the hidden thinking tokens. Some providers discount cached input or batch requests.
This site shows a blended price that assumes three input tokens for every output token, so different models can be compared on one number.
Related ranking: Cheapest models
- 1.Mistral Nemo Mistral AI$0.022
- 2.Ling 3.0 Flash VL inclusionAI (Ant Group)$0.031
- 3.Ling 3.0 Flash inclusionAI (Ant Group)$0.032
- 4.gpt-oss-20b OpenAI$0.036
- 5.Granite 4.0 Micro IBM$0.041
Ranked by blended API price (75% input + 25% output, USD per 1M tokens). Free promotional listings excluded. Source: OpenRouter. Full ranking
Related terms
Explanation written by AI Stats Live editors; last reviewed 29 Sep 2026. Examples and rankings update automatically from the sources listed on the methodology page.