Cost of Using Shared Models

We offer a range of shared models that enable users to quickly engage in chat conversations on ConsoleX and run evaluations on EvalsOne without having to add their own model integrations.

When using shared models, you will need to consume your Credits on EvalsOne/ConsoleX.

Understanding the cost of different models is crucial for managing your credits effectively. Below is our current pricing structure:

Cost of shared models

Model	Usage	Cost
GPT-4o-mini	1M input tokens	$0.195
	1M output tokens	$0.78
GPT-4o (The newest gpt-4o-2024-08-06 model)	1M input tokens	$3.25
	1M output tokens	$13
GPT-4 Turbo	1M input tokens	$13.00
	1M output tokens	$39.00
GPT-4	1M input tokens	$39
	1M output tokens	$78
Claude 3.5 Sonnet	1M input tokens	$3.90
	1M output tokens	$19.50
Claude 3 Haiku	1M input tokens	$0.33
	1M output tokens	$1.63
Claude 3 Opus	1M input tokens	$19.5
	1M output tokens	$97.50

Note: Prices are in USD and are subject to change. Always refer to the latest pricing information on our platform.

The prices for the shared models we offer are typically 1.3 times the official model prices to ensure the sustainability of our services. We reserve the right to adjust our pricing strategy within a reasonable range in the future.

Understanding the Table:

• Context: The maximum number of tokens the model can process in a single call.
• Input/1k Tokens: Cost per 1,000 tokens of input text.
• Output/1k Tokens: Cost per 1,000 tokens of generated text.
• Per Call: Total cost per API call (Input + Output).
• Total: Cost per 1,000,000 tokens.

Tips for Cost-Effective Usage:

• For shorter, simpler tasks, consider using models with lower costs like GPT-3.5 Turbo or Claude 3 Haiku.
• For complex tasks requiring advanced capabilities, GPT-4o or Claude 3.5 Sonnet might provide better value despite higher costs.
• Monitor your usage and adjust your choice of model based on your specific needs and budget.