The cost to use a AI depends on the usage you are : number of queries, size of the input context and the output generated.
If you use a local model, it mainly depends on your electricity consomption and the material invest.
If you use an API, the cost is specific to the provider and the model used.
Here a table of principal providers pricing
# OpenAI
Pricing : https://openai.com/pricing
Rate limits : https://platform.openai.com/docs/models
# MistralAI
Pricing & Rate limits : https://docs.mistral.ai/platform/pricing/
# Hugging Face
Pricing : https://huggingface.co/pricing
Rate limits : https://huggingface.co/docs/api-inference/faq#rate-limits
# Anthropic
Pricing : https://www.anthropic.com/pricing
Rate limits : https://docs.anthropic.com/claude/reference/errors-and-rate-limits
# Google Generative AI
Pricing : https://cloud.google.com/vertex-ai/docs/generative-ai/pricing
Rate limits : https://cloud.google.com/generative-ai-app-builder/quotas