i need help with seeing what the agent pricing is
#agent pricing
surreal phoenix
narrow marshBOT
Hi @surreal phoenix. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
The pricing for using an agent (also referred to as an assistant) with OpenAI involves several components:
-
Assistant Retrieval Cost: You would pay $0.02 per assistant per day. If you have just one assistant, it would cost you $0.02 per day [source (https://community.openai.com/t/637726)].
-
Token Costs: The tokens used for the Assistant API are billed at the chosen language model’s per-token input/output rates. The assistant intelligently chooses which context from the thread to include when calling the model. The exact quantity of tokens used cannot be predicted as it depends on your prompts, replies, and the tokens used by the retrieval system to complete each request. This could potentially be up to the context limit of 128k tokens [source (https://community.openai.com/t/637726), source (https://community.openai.com/t/553660)]. The cost per token depends on the specific GPT model you are using. For example, for GPT-4 with a context length of 128k (e.g.,
gpt-4-turbo), the price is $10.00 per 1 million prompt tokens and $30.00 per 1 million sampled tokens [source (https://help.openai.com/en/articles/7127956-how-much-does-gpt-4-cost)]. -
File Search Cost: If you're using the
file_searchtool, it costs $0.10 per GB of vector-storage per day. However, the first GB is free [source (https://community.openai.com/t/724092)]. -
Code Interpreter Cost: If your assistant uses the Code Interpreter, it is priced at $0.03 per session. Each session is active by default for one hour [source (https://help.openai.com/en/articles/8550641-assistants-api-v2-faq#h_c2c7089a6f)].
Please note that you will be charged for the whole thread, the retrieved documents, your new query, and the new output every time the thread runs [source (https://community.openai.com/t/606285)].
You can monitor your daily costs, your monthly spend, and the credit grants for the current and previous months in the "View by cost" section of your usage dashboard [source (https://help.openai.com/en/articles/8554956-understanding-the-usage-dashboard#h_90d1c4aa37)].
For more detailed information, please refer to the OpenAI Pricing page (https://openai.com/pricing).
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
radiant grove
@narrow marsh how do you predict the next ai model costs? how have been pricing doing upto date?
narrow marshBOT
Let me check the knowledge base for you again. One moment! :books:
Predicting the cost of future AI models is a complex task and depends on a variety of factors, including the complexity of the model, the resources required to train and run it, and market conditions. OpenAI does not provide a specific method for predicting future model costs.
However, you can get an idea of how pricing has evolved by looking at the deprecation history and the release notes of various models. For example, the deprecation history shows that older models like gpt-4-0314, gpt-3.5-turbo-0301, and gpt-4-32k-0314 were replaced by newer models like gpt-4o and gpt-3.5-turbo [source (https://platform.openai.com/docs/deprecations)]. The pricing for these models varied, with costs per 1M input tokens and 1M output tokens ranging from $15.00 to $60.00 and $20.00 to $120.00 respectively.
The release notes also provide information about the introduction of new models like GPT-4o and their features [source (https://help.openai.com/en/articles/4936856-what-are-tokens-and-how-to-count-them#h_4f7e925c57)]. The pricing for GPT-4, for example, varies depending on the context length of the model, with costs per 1M prompt tokens and 1M sampled tokens ranging from $10.00 to $60.00 and $30.00 to $120.00 respectively [source (https://help.openai.com/en/articles/7127956-how-much-does-gpt-4-cost)].