AI model rankings.
Every list below is recomputed from live prices across 212 providers, every day (September 2026 edition). Prices are per million tokens, three parts input to one part output. A price below the maker's own with nothing published to account for it is listed on the model's page and ranked nowhere.
Cheapest LLM API
#1 today: Llama 3.2 1B Instruct at Inference
from $0.04/M tokensview →
Cheapest models with 128k context
#1 today: Llama 3.2 3B Instruct at Inference
from $0.08/M tokensview →
Cheapest open-weight models
#1 today: Llama 3.2 1B Instruct at Inference
from $0.04/M tokensview →
Cheapest models with tool calling
#1 today: Mistral Nemo at Nous Portal
from $0.08/M tokensview →
Cheapest models that read images
#1 today: GLM-5.3-Flash at Merge Gateway
from $0.10/M tokensview →
Cheapest reasoning models
#1 today: GLM-5.3-Flash at Merge Gateway
from $0.10/M tokensview →
Biggest savings against the list price
#1 today: Nano Banana at Kilo Gateway
from $1.70/M tokensview →
Free tiers worth using
#1 today: Nano Banana at Kilo Gateway
from $1.70/M tokensview →